Thirty-Third Conference on Artificial Intelligence
January 27 – 28, 2019
Honolulu, Hawaii, USA
What Is the Tutorial Forum?
The Tutorial Forum provides an opportunity for researchers and practitioners to spend two days each year exploring exciting advances in disciplines outside their normal focus. We believe this type of forum is essential for the cross-fertilization, cohesiveness, and vitality of the AI field. We all have a lot to learn from each other; the Tutorial Forum promotes the continuing education of each member of AAAI.
Schedule
Half-day tutorials are 4 hours, including breaks; quarter-day tutorials are one hour and 45 minutes with no break. Quarter-day tutorials are denoted by a ‘Q’ at the end of the tutorial code.
Sunday, January 27, 2019
8:30 AM – 12:30 PM
- SA1: Answer Set Engineering
Tutorial Materials: (available after January 5) - SA2: Deep Multi-view Visual Data Analytics
Tutorial Materials: (available after January 5) - SA3: Deep Reinforcement Learning with Applications in Transportation
Tutorial Materials: (available after January 5) - SA4: On Explainable AI: From Theory to Motivation, Applications and Limitations
Tutorial Materials: (available after January 5) - SA5: Plan, Activity and Intent Recognition (PAIR)
Tutorial Materials: (available after January 5)
1:30 – 5:30 PM
- SP1: Behavior Analytics: Methods and Applications
Tutorial Materials: (available after January 5) - SP2: Building Deep Learning Applications for Big Data Platforms
Tutorial Materials: https://jason-dai.github.io/aaai2019/ - SP3: New Frontiers of Automated Mechanism Design for Pricing and Auctions
Tutorial Materials: https://sites.google.com/view/amdtutorial/home
1:30 – 3:15 PM
- SP4Q: Federated Learning: User Privacy, Data Security and Confidentiality in Machine Learning
Tutorial Materials: https://www.fedai.org/#/conferences/link_aaai2019 - SP5Q: An Overview of the International Planning Competition
Tutorial Materials: https://www.nms.kcl.ac.uk/andrew.coles/PlanningCompetitionAAAISlides.html - SP6Q: Presenting a Paper
Tutorial Materials: https://wp.me/P3qAAw-76
3:45 – 5:15 PM
- SP7Q: Planning and Scheduling Approaches for Urban Traffic Control
Tutorial Materials: https://helios.hud.ac.uk/scommv/storage/TutorialSlides.pdf - SP8Q: The Road to Industry
Tutorial Materials: (available after January 5)
Monday, January 28, 2019
8:30 AM – 12:30 PM
- MA1: Adversarial Machine Learning
Tutorial Materials: https://aaai19adversarial.github.io/index.html#org - MA2: Deep Bayesian and Sequential Learning
Tutorial Materials: http://chien.cm.nctu.edu.tw/home/aaai-tutorial/ - MA3: Multi-Agent Pathfinding: Models, Solvers, and Systems
Tutorial Materials: http://ktiml.mff.cuni.cz/~bartak/AAAI2019/ - MA4: Neural Vector Representations beyond Words: Sentence and Document Embeddings
Tutorial Materials: http://gerard.demelo.org/teaching/embedding-tutorial/ - MA5: Recent Advances in Scalable Retrieval of Personalized Recommendations
Tutorial Materials: https://preferred.ai/aaai19-tutorial/
1:30 – 5:30 PM
- MP1: End-to-end Goal-oriented Question Answering Systems
Tutorial Materials: https://www.slideshare.net/QiHe2/aaai-2019-tutorial-endtoend-goaloriented-question-answering-systems - MP2: Graph Representation Learning
Tutorial Materials: (available after January 5) - MP3: Imagination Science: Beyond Data Science
Tutorial Materials: https://people.cs.umass.edu/~mahadeva/AAAI_2019_Tutorial/Welcome.html - MP4: Integrating Human Factors into AI for Fake News Prevention: Challenges and Opportunities
Tutorial Materials: (available after January 5)
1:30 PM – 3:15 PM
- MP5Q: Knowledge-based Sequential Decision-Making under Uncertainty
Tutorial Materials: http://www.cs.binghamton.edu/~szhang/2019_aaai_tutorial/
3:45 – 5:15 PM
- MP6Q: Human Identification at a Distance by Gait Recognition
Tutorial Materials: http://yushiqi.cn/research/aaai19-gait-recognition-tutorial
SA1: Answer Set Engineering
Roland Kaminski, Javier Romero, Torsten Schaub, Philipp Wanko
Answer Set Programming (ASP) has become an established paradigm for Knowledge Representation and Reasoning, in particular, when it comes to solving knowledge-intense combinatorial (optimization) problems. ASP’s unique pairing of a simple yet rich modeling language with highly performant solving technology has led to an increasing interest in ASP systems in academia as well as industry. To further boost this development and make ASP fit for real-world applications, it is indispensable to possess means for easy integration into software environments and for adding complementary forms of reasoning. These means are furnished by multi-shot and theory reasoning in modern ASP systems.
The participant will learn (i) how to use multi-shot solving to integrate ASP systems into software environments and to deal with solver objects to model complex reasoning modes (see below) and (ii) how to use theory reasoning to integrate dedicated reasoners into ASP systems and to extend the modeling language to express corresponding constraints. Or, in a nutshell, how to exploit modern ASP technology for solving complex reasoning problems. This will be detailed by means of the ASP system CLINGO and its application programming interface.
Roland Kaminski
University of Potsdam
Roland Kaminski is a research engineer at the University of Potsdam, Germany. His scientific interests are focused on the implementation of ASP systems. He is the main developer of the widely used ASP system CLINGO. Recently, his focus is the improvement of CLINGO’s application programming interface, providing the basis for systems extending core ASP solving, for example, ASPRIN and CLINGO[DL].
Javier Romero
University of Potsdam
Javier Romero is a researcher at the University of Potsdam, Germany. He received a MSc. in Computer Science from the University of Corunna in 2006 and in Philosophy from the University of Santiago de Compostela in 2008, both in Spain. His research interests are knowledge representation and logic programming, and his research focuses on extensions of ASP for declarative heuristics and preference reasoning. He is a member of the open-source project potassco.org developed at Potsdam and the main developer of ASPRIN.
Torsten Schaub
University of Potsdam
Torsten Schaub is University Professor for knowledge processing and information systems at the University of Potsdam. Torsten is a fellow of the European Association for Artificial Intelligence EurAI and serves as president of the Association of Logic Programming. His research interest range from the theoretic foundations to the practical implementation of reasoning from incomplete, inconsistent, and evolving information. His research focus lies on ASP and materializes at potassco.org, the home of the open source project Potassco bundling software for ASP developed at the University of Potsdam.
Philipp Wanko
University of Potsdam
Philipp Wanko is a doctoral student and researcher at the University of Potsdam, Germany. His interests are focused on hybrid reasoning and multi-objective optimization using ASP. Both fields are of particular interest to his main field of application, Design Space Exploration, where implementations of embedded systems featuring a complex network-like hardware structure are automatically synthesized. To effectively encode timing constraints featured in this application, he developed the system CLINGO[DL] that combines ASP with different constraints. Furthermore, he extended CLINGO[DL] with the capacity for multi-objective optimization, in particular, enumerating Pareto-optimal (non-dominated) solutions.
SA2: Deep Multi-View Visual Data Analytics
Zhengming Ding, Hongfu Liu, Handong Zhao
Multi-view data are extensively accessible nowadays thanks to various types of features, viewpoints, and different sensors. For example, the most popular commercial depth sensor Kinect uses both visible light and near-infrared sensors for depth estimation; automatic driving uses both visual and radar sensors to produce real-time 3D information on the road; and face analysis algorithms prefer face images from different views for high-fidelity reconstruction and recognition. All of them tend to facilitate better data representation in different application scenarios. Recently there are a bunch of approaches proposed to deal with multi-view visual data. Our tutorial covers the most popular deep multi-view learning algorithms proposed recently, centered around four major applications, i.e., multi-view clustering, multi-view classification, zero-shot learning, and domain adaptation. By participating in the tutorial, the audience will gain a broad knowledge of multi-view learning, including its most recent advance in visual data analysis and detailed analysis of typical algorithms/frameworks. In addition, the audience will walk through a variety of popular visual data analysis tools based on deep learning. Specifically, only basic knowledge regarding classification and clustering is required, as we will go over other related algorithms together with the multi-view problem setting at the beginning.
Zhengming Ding
Indiana University-Purdue University Indianapolis
Zhengming Ding is a tenure-tracked faculty in the Department of Computer Information and Technology, Indiana University-Purdue University Indianapolis. He received Ph.D. from Northeastern University, USA. His research includes transfer learning, multi-view learning, and deep learning. He received the best paper award (SPIE 2016) and a best paper candidate (ACM MM 2017).
Hongfu Liu
Brandeis University
Hongfu Liu is currently a tenure-track assistant professor at Michtom School of Computer Science at Brandeis University. He received his Ph.D. in the Department of Electrical & Computer Engineering at Northeastern University. Before joining NEU, He got his master’s and bachelor’s degrees and majored in management at Beihang University.
Handong Zhao
Adobe Research
Handong Zhao is a research scientist at Adobe Research, San Jose, CA. He earned his Ph.D. degree from the Department of Electrical and Computer Engineering, Northeastern University, Boston, MA. His current research interest includes machine learning and its applications to multimedia content analysis and marketing analytics.
SA3: Deep Reinforcement Learning with Applications in Transportation
Zhiwei Qin, Jian Tang, Jieping Ye
Transportation, particularly the mobile ride-sharing domain, has a number of traditionally challenging dynamic decision problems that have long threads of research literature and readily stand to benefit tremendously from artificial intelligence (AI). Some core examples include online ride order dispatching, route planning, and traffic signal control.
Reinforcement learning (RL) is a machine learning paradigm that trains an agent to learn to take optimal actions (as measured by the total cumulative reward achieved) in an environment through interactions with it and getting feedback signals. Thanks to the rapid advancement in deep learning research and computing capabilities, the integration of deep neural networks and RL has generated explosive progress in the latter for solving complex, large-scale learning problems, attracting a huge amount of renewed interest in recent years. It presents a tremendous potential to solve some hard problems in transportation in an unprecedented way.
This tutorial is targeted to researchers and practitioners with a general machine learning background and interested in working on applications of deep RL (DRL) in transportation. The goal of this tutorial is to provide the audience with a guided introduction to this exciting area of AI with specially curated application case studies in transportation. The prerequisite knowledge assumed of the audience includes a basic understanding of deep neural networks, optimization, and machine learning concepts. Exposure to Markov decision process and operations research, in general, is preferred.
Zhiwei (Tony) Qin
DiDi AI Labs
Tony Qin leads the reinforcement learning research at DiDi AI Labs. He received his Ph.D. in Operations Research from Columbia University. His research interest lies broadly in operations research, machine learning/deep learning, and big data analytics, with applications in smart transportation and E-commerce.
Jian Tang
DiDi AI Labs
Jian Tang is the chief scientist of intelligent control at DiDi AI Labs. His research interests include Machine Learning, Big Data, and Wireless Networking. He has published over 130 papers in premier journals and conferences. He received an NSF CAREER award in 2009.
Jieping Ye
DiDi AI Labs
Jieping Ye is head of DiDi AI Labs, and a VP of Didi Chuxing. He is also an associate professor at University of Michigan, Ann Arbor. His research interests include big data, machine learning, and data mining with applications in transportation and biomedicine.
SA4: On Explainable AI: From Theory to Motivation, Applications, and Limitations
Luca Costabello, Freddy Lecue, Fosca Giannotti, Riccardo Guidotti, Pascal Hitzler, Pasquale Minervini, Kamruzzaman Sarker
The future of AI lies in enabling people to collaborate with machines to solve complex problems. Like any efficient collaboration, this requires good communication, trust, clarity, and understanding. XAI (eXplainable AI) addresses such challenges by combining symbolic AI and traditional machine learning. Such a topic has been studied for years by different AI communities, leading to different definitions, evaluation protocols, motivations, and results. This tutorial is a snapshot of XAI to date and surveys the work achieved by the AI community, with a focus on machine learning and symbolic AI. We will motivate the needs of XAI in real-world and large-scale applications while presenting state-of-the-art techniques and best practices.
In the first part of the tutorial, we introduce the different aspects of explanations in AI. We then focus on two specific approaches: (i) XAI using machine learning and (ii) XAI using a combination of graph-based knowledge representation and machine learning. For both, we get into the specifics of the approach, the state of the art, and the research challenges for the next steps. The final part of the tutorial gives an overview of real-world applications of XAI.
Luca Costabello
Accenture Labs Dublin
Luca Costabello is a research scientist at Accenture Labs Dublin. His research interests span knowledge graphs applications, machine learning for graphs, and explainable AI. Before joining Accenture, Luca was a research scientist at Fujitsu Ireland, PhD student at Inria France, and a research engineer at Telecom Italia.
Fosca Giannotti
National Research Council, Pisa, Italy
Fosca Giannotti is the Director of Research at the Information Science and Technology Institute “A. Faedo” of the National Research Council, Pisa, Italy. Fosca Giannotti is a scientist in Data Mining, Machine Learning, and Big Data Analytics. Fosca leads the Pisa KDD Lab – Knowledge Discovery and Data Mining Laboratory http://kdd.isti.cnr.it, a joint research initiative of the University of Pisa and ISTI-CNR, founded in 1994 as one of the earliest research labs centered on data mining. Fosca’s research focus is on social mining from big data: human dynamics, social networks, diffusion of innovation, privacy-enhancing technology, and explainable AI. She has coordinated tens of research projects and industrial collaborations. Fosca is now the coordinator of SoBigData, the European research infrastructure on Big Data Analytics and Social Mining, an ecosystem of ten cutting-edge European research centres providing an open platform for interdisciplinary data science and data-driven innovation http://www.sobigdata.eu. From 2012-2015 Fosca has been general chair of the Steering Board of ECML-PKDD (European Conference on Machine Learning) and is currently a member of the steering committee EuADS (European Association on Data Science) and of the AIIS: Italian Lab. of Artificial Intelligence and Autonomous Systems.
Riccardo Guidotti
University of Pisa, Italy
Riccardo Guidotti is a post-doc researcher at the Department of Computer Science University of Pisa, Italy, and a member of the KDDLab, a joint research group with the ISTI-CNR in Pisa. Riccardo Guidotti was born in 1988 in Pitigliano (GR) Italy. In 2013 and 2010, he graduated cum laude in Computer Science (MS and BS) at the University of Pisa. He received a Ph.D. in Computer Science with a thesis on Personal Data Analytics from the same institution. He won the IBM fellowship program and was an intern at IBM Research in Dublin, Ireland in 2015. He recently won the Next Generation Data Scientist Award at DSAA 2018. His research interests are in personal data mining, clustering, explainable models, and analysis of transactional data related to recipes and to migration flows.
Pascal Hitzler
Wright State University
Pascal Hitzler is endowed NCR Distinguished Professor, Brage Golding Distinguished Professor of Research, and Director of Data Science at the Department of Computer Science and Engineering at Wright State University in Dayton, Ohio, U.S.A. He is director of the Data Semantics (DaSe) Lab and a founding member of the Data Science and Security Cluster (DSSC). From 2004 to 2009, he was Akademischer Rat at the Institute for Applied Informatics and Formal Description Methods (AIFB) at the University of Karlsruhe in Germany, and from 2001 to 2004, he was a postdoctoral researcher at the Artificial Intelligence Institute at TU Dresden in Germany. In 2001 he obtained a Ph.D. in Mathematics from the National University of Ireland, University College Cork, and in 1998 a Diplom (Master equivalent) in Mathematics from the University of Tübingen in Germany. His research record lists over 400 publications in such diverse areas as semantic web, artificial intelligence, neural-symbolic integration, knowledge representation and reasoning, machine learning, denotational semantics, and set-theoretic topology. His research is highly cited. He is the founding Editor-in-chief of the Semantic Web journal, the leading journal in the field, and of the IOS Press book series Studies on the Semantic Web. He is co-author of the W3C Recommendation OWL 2 Primer, and of the book Foundations of Semantic Web Technologies by CRC Press, 2010, which was named as one out of seven Outstanding Academic Titles 2010 in Information and Computer Science by the American Library Association’s Choice Magazine, and has translations into German and Chinese. He is on the editorial board of several journals and book series and a founding steering committee member of the Neural-Symbolic Learning and Reasoning Association and the Association for Ontology Design and Patterns, and he frequently acts as conference chair in various functions, including e.g., General Chair (ESWC2019, US2TS2018), Program Chair (FOIS 2018, AIMSA2014), Track Chair (ISWC2018, ESWC2018, ISWC2017, ISWC2016, AAAI-15), Workshop Chair (K-Cap2013), Sponsor Chair (ISWC2013, RR2009, ESWC2009), Ph.D. Symposium Chair (ESWC 2017). For more information about him, see http://www.pascal-hitzler.de.
Sarker Kamruzzaman
Wright State University
Md Kamruzzaman Sarker is a Ph.D. student at Wright State University. His current research focuses on making machine learning algorithms’ decision making more transparent. He is also interested in making ontology engineering processes easier and more human-friendly. Before his Ph.D., he worked at Samsung Electronics as a software engineer.
Freddy Lecue
Accenture Labs Dublin and Inria, France
Dr. Freddy Lecue (Ph.D. 2008, Habilitation 2015) is an Artificial Intelligence R&D lead at Accenture Labs – Ireland and research associate at INRIA – France. His research area is at the frontier of intelligent learning/reasoning systems. He has a strong interest in Explainable AI, i.e., AI systems, models, and results that can be explained to human and business experts.
Pasquale Minervini
University College London
Pasquale is a Research Associate at University College London (UCL), United Kingdom. He received a Ph.D. in Computer Science from the University of Bari, Italy, with a thesis titled “Mining Methods for the Web of Data”. After obtaining his Ph.D., he worked as a postdoctoral researcher at the University of Bari, Italy, and at the INSIGHT Centre for Data Analytics (INSIGHT), Galway, Ireland. At INSIGHT, he worked in the Knowledge Engineering and Discovery (KEDI) group, composed of researchers and engineers from INSIGHT and Fujitsu Ireland Research and Innovation. Over the course of his research career, Pasquale published peer-reviewed papers in top-tier AI conferences (such as UAI, AAAI, ICDM, CoNLL, ECML, and ESWC), receiving two best paper awards. He is the main inventor of a patent application assigned to Fujitsu Ltd. For more information about him, see http://www.neuralnoise.com.
SA5: Plan, Activity, and Intent Recognition (PAIR)
Sarah Keren, Reuth Mirsky, Christopher Geib
Plan, activity, intent, and goal recognition all involve making inferences about other actors (software agents, robots, or humans) from observations of their behavior, i.e., their interaction with the environment and with each other. This synergistic area of research combines techniques from user modeling, machine vision, automated planning, intelligent user interfaces, human-computer interaction, autonomous and multi-agent systems, natural language understanding, and machine learning. It plays a crucial role in a wide variety of applications including assistive technology, software assistants, computer and network security, behavior recognition, human-robot collaboration, and more. This widespread diversity of applications and disciplines, while producing a wealth of ideas, models, tools, and results, has contributed to fragmentation in the field. This tutorial seeks to amend this by providing an overview of the recognition literature, as well as a description of the core elements that comprise a recognition problem. In particular, using several motivating examples, we will describe the different recognition tasks, outline their scope, and describe the relationship between them. Finally, we will describe the state of the art of recognition research, highlighting the leading computational representations, algorithms, empirical methodologies, and applications. This half-day tutorial is aimed at general AI students and researchers that wish to explore potential research directions in recognition or use recognition to enhance their ongoing research.
Sarah Keren
Harvard University
Sarah Keren is a post-doctoral fellow at Harvard University. She obtained his Ph.D. from the Technion – Israel Institute of Technology. Sarah’s research focuses on redesigning environments for optimized utility. In particular, her Ph.D. established the problem of Goal Recognition Design, dedicated to redesigning goal recognition setting for facilitating online goal recognition.
Reuth Mirsky
Ben-Gurion University
Reuth Mirsky is a Ph.D. candidate at Ben-Gurion University. Her adviser is Dr. Kobi Gal. Reuth’s research focuses on plan recognition in real-world environments. In particular, her Ph.D. focused on plan recognition challenges, such as compact problem representation, efficient domain design and hypotheses disambiguation. Her algorithms have been applied in various tasks for education, clinical treatment, and finance.
Christopher Geib
SIFT LLC
Christopher Geib is a Principal Researcher at SIFT LLC. He has been the principal architect of multiple probabilistic plan recognition systems, including the ELEXIR system, which has demonstrated state of the art plan recognition and planning capabilities based on a single shared and learnable representation of the domain.
SP1: Behavior Analytics: Methods and Applications
Longbing Cao
Complex behaviors are widely seen in artificial and natural intelligent systems, on the internet, physical and virtual systems, social and online networks, multi-agent systems, and brain systems. The in-depth understanding of complex behaviors has been increasingly recognized as a crucial means for disclosing interior driving forces, causes, and impact on businesses in handling many challenging issues. However, traditional behavior modeling mainly relies on qualitative methods from behavioral science and social science perspectives. The so-called behavior analysis in data analytics and learning often focuses on human demographic and business usage data, in which behavior-oriented elements are hidden in routinely collected transactional data. As a result, it is ineffective or even impossible to deeply scrutinize native behavior intention, lifecycle, dynamics, and impact on complex problems and business issues. In this tutorial, we will present an overview of behavior analytics, review and discuss the state-of-the-art and newly emerged techniques for complex behavior analytics, which cover high-impact behavior sequence analysis, impact-oriented combined behavior analysis, high utility behavior analysis, nonoccurring behavior analysis, coupled/group/collective behavior analysis, statistical modeling of coupled behaviors, probabilistic modeling of sparse rating behaviors, understanding behavior choice and attraction, behavior analysis with recurrent networks, behavior analysis in visual data, behavior learning from demonstrations. We will show that in-depth behavior analytics creates new opportunities, directions, and means for learning and analysis of complex behaviors in both physical and virtual organizations.
For more information, please see http://203.170.84.89/~idawis33/DataScienceLab/news/aaai2019-tutorial-behavior-analytics-methods-and-applications/.
Longbing Cao
UTS
Professor Longbing Cao holds a Ph.D. in Pattern Recognition and Intelligent Systems from Chinese Academy of Sciences and another Ph.D. in Computing Science at UTS. He has published some 300 publications, four monographs, and four edited books in recent 15 years. He has been working on data science and analytics research, education, development, and enterprise applications since he was a CTO and then joined UTS. Motivated by real-world significant and common challenges, he has been leading the team to develop theories, tools, and applications for new areas, including non-IID learning, actionable knowledge discovery, behavior informatics, and complex intelligent systems, in addition to issues generally concerned with artificial intelligence, knowledge discovery, machine learning, and their enterprise applications. In data science and analytics, he initiated the Data Science and Knowledge Discovery lab at UTS in 2007, the Advanced Analytics Institute in 2011, and the degrees Master of Analytics (Research) and Ph.D. in Analytics in 2011, which are recognized as the world’s first degrees in data science, the IEEE Task Force on Data Science and Advanced Analytics (DSAA) and IEEE Task Force on Behavior, Economic and Soci-cultural Computing in 2013, the IEEE Conference on Data Science and Advanced Analytics (DSAA), the ACM SIGKDD Australia and New Zealand Chapter in 2014, and the International Journal of Data Science and Analytics with Springer in 2015. He served as program and general chair of conferences such as KDD2015. In enterprise data science innovation, his team has successfully delivered many large projects for government and business organizations in over ten domains, including finance/capital markets, banking, health and car insurance, health, telco, recommendation, online business, education, and the public sector including ATO, DFS, DHS, DIBP, and IP Australia, resulting in billions of dollar savings and mentions in government, industry, media and OECD reports. In 2013, AAI was the only organization specially mentioned in the Governments first big data paper: Big Data Strategy Issues Paper. He has been delivered invited and keynote speeches to over 20 conferences, guest lectures and seminars to many universities, and tutorials to conferences including AAAI, IJCAI and KDD.
SP2: Build Deep Learning Applications for Big Data Platforms
Jason Dai
Recent breakthroughs in artificial intelligence applications have brought deep learning to the forefront of new generations of data analytics. In this tutorial, we will present the practice and design tradeoffs for building large-scale deep learning applications (such as computer vision and NLP) for production data and workflow on Big Data platforms. We will provide an overview of emerging deep learning frameworks for Big Data (e.g., BigDL, TensorFlowOnSpark, Deep Learning Pipelines for Apache Spark, etc.) and present the underlying distributed systems and algorithms. More importantly, we will show how to build and productionize end-to-end deep learning application pipelines for Big Data (on top of Analytics Zoo, a unified analytics + AI platform for distributed TensorFlow, Keras, and BigDL on Apache Spark), using real-world use cases (such as Azure, JD.com, World Bank, Midea/KUKA, etc.)
Jason Dai
Intel
Jason Dai is a senior principal engineer and CTO of Big Data Technologies at Intel, responsible for leading the global engineering teams (in both Silicon Valley and Shanghai) on the development of advanced big data analytics and machine learning. He is the program co-chair of the O’Reilly AI Conference in Beijing, a founding committer and PMC member of Apache Spark, and the creator of BigDL.
SP3: New Frontiers of Automated Mechanism Design for Pricing and Auctions
Maria-Florina Balcan, Tuomas Sandholm, Ellen Vitercik
Automated mechanism design is a powerful and prominent approach to pricing and auction design, a field of game theory with significant real-world impact. This line of work uses optimization and machine learning to devise selling mechanisms based on data. By automating mechanism design, we overcome obstacles faced by traditional, manual techniques, which have been stuck for decades due to inherent computational complexity challenges. Remarkably, the revenue-maximizing mechanism is not known even for just two items for sale! In this tutorial, we cover the rapidly growing area of automated mechanism design for revenue maximization.
Maria-Florina Balcan
Carnegie Mellon University
Maria-Florina Balcan is an Associate Professor of Computer Science at Carnegie Mellon University, working in machine learning, game theory, and algorithms. She was Program Committee Co-chair for COLT 2014 and ICML 2016. She is currently an IMLS board member, Tutorial Chair for ICML 2019, and Workshop Chair for FOCS 2019.
Tuomas Sandholm
Carnegie Mellon University
Tuomas Sandholm is Angel Jordan Professor of Computer Science at Carnegie Mellon University and Co-Director of CMU AI. He is the Founder and Director of the Electronic Marketplaces Laboratory. He is a successful serial entrepreneur. He has fielded over 800 combinatorial auctions, worth over $60 billion. He is Founder and CEO of Optimized Markets, Strategic Machine, and Strategy Robot.
Ellen Vitercik
Carnegie Mellon University
Ellen Vitercik is a Ph.D. student at Carnegie Mellon University. Her primary research interests are artificial intelligence, machine learning, theoretical computer science, and the interface between economics and computation. Her honors include a National Science Foundation Graduate Research Fellowship.
SP4Q: Federated Learning: User Privacy, Data Security and Confidentiality in Machine Learning
Yang Liu, Qiang Yang, Zhuoshi Wei, Tianjian Chen
Today’s AI faces two major challenges. One is that in most industries, data exists in the form of isolated islands. The other is the increasing demand for AI to be aware of user privacy and data security. We give an overview of these challenges and survey recent works on secure federated learning to meet them. We will describe the federated learning framework by considering horizontal federated learning, vertical federated learning, and federated transfer learning. We provide definitions, architectures, and applications for the federated learning framework. We also survey related works in AI security, privacy and confidentiality. We will show some typical application scenarios of the technology. Finally, we describe a development roadmap for federated and transfer learning.
Yang Liu
WeBank
Yang Liu is a Senior Researcher in AI Department of WeBank, China. Her research interests include machine learning, federated learning, federated transfer learning, and applications of these technologies in the financial industry. She received her Ph.D. in Chemical and Biological Engineering from Princeton University in 2012 and her Bachelor’s degree from Tsinghua University.
Qiang Yang
Hong Kong University of Science and Technology
Qiang Yang is a Chair Professor in Department of Computer Science and Engineering at the Hong Kong University of Science and Technology. Qiang Yang received his Ph.D. from the Computer Science Department of the University of Maryland, College Park, in 1989. He is a Fellow of AAAI, ACM, IEEE, AAAS, and IAPR. His professional services include being the IJCAI President (2017-2019), ACM SIGART (SIGAI) Vice Chair (2009-2012); PC Chair of IJCAI (2015), ACM KDD (2010); Conference Chair of ACM KDD (2012). He is the Editor in Chief of IEEE Transactions on Big Data.
Zhuoshi Wei
WeBank
Zhuoshi Wei is a Researcher in the Department of AI at WeBank, China. Zhuoshi Wei received her B.S. from the School of Electronic and Information Engineering, Beijing Jiaotong University, in 2003 and her Ph.D. in Pattern Recognition and Intelligent Systems from the National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Sciences in 2009. She had been a postdoctoral researcher at the Department of Radiology and Imaging Sciences, National Institutes of Health Clinical Center, USA (2010-2013). She was a Principal Data Scientist at Capital One Financial Corporation (2015-2017). She joined the Department of AI at WeBank in 2018.
Tianjian Chen
WeBank
Tianjian Chen is the Deputy General Manager of the Department of AI at WeBank, China. Tianjian Chen received his Bachelor’s degree in Electronic Engineering from Tsinghua University in 2006. He had been a Software Engineer (2006-2008) and Senior Software Engineer (2008-2009) at Baidu, System Architect at Xunlei Networking Technologies (2009), System Architect at BGI (2010-2011), Senior Architect (2011-2014) and Principal Architect (2014-2018) at Baidu. He joined the Department of AI at WeBank in 2018.
SP5Q: An Overview of the International Planning Competition
Amanda Coles, Andrew Coles, Florian Pommerening, Álvaro Torralba
AI Planning is one of the oldest sub-areas in AI. The International Planning Competition (IPC) is organized in the context of the International Conference on Planning and Scheduling (ICAPS). It empirically evaluates state-of-the-art planning systems on a number of benchmark problems. The goals of the IPC are to promote planning research, highlight challenges in the planning community, and provide new and interesting problems as benchmarks for future research. The IPC has an important role in the ICAPS community, being a forum to compare different algorithmic ideas and implementations, and setting standards for research and evaluation in the area.
The goal of this tutorial is to inform participants about the IPC series, specifically, we focus on the deterministic and temporal tracks, of which we were the organizers for the most recent competition. We aim to equip participants with the knowledge of the current state-of-the-art and the key open challenges of classical and temporal planning, as well as familiarize them with the tools and resources available to help those developing planners.
Amanda Coles
King’s College London
Amanda Coles is a senior lecturer in the Department of Informatics, King’s College London, specialising in mixed discrete-continuous (hybrid) planning. She is a member of the ICAPS executive council; participant in three past International Planning Competitions; and organiser of the IPC Learning Track in 2011 and Temporal Track in 2018
Andrew Coles
King’s College London
Andrew Coles is a senior lecturer in the Department of Informatics at King’s College London. He is a principal investigator at King’s for the ERGO and ADE projects, applying temporal planning to problems in the area of autonomous space robotics. He was organiser of the International Planning Competition tracks in 2011 and 2018.
Florian Pommerening
University of Basel
Florian Pommerening is a postdoctoral researcher in the AI group at the University of Basel, Switzerland, where he completed his Ph.D. from 2012 to 2017. His main research interest is classical automated planning. Together with Álvaro Torralba, he organized the classical tracks of the IPC 2018.
Álvaro Torralba
Saarland University
Álvaro Torralba is a postdoctoral researcher in the FAI group at Saarland University. He received his Ph.D. from Universidad Carlos III de Madrid in 2015. His main research interests are heuristic search and automated planning. Together with Florian Pommerening, he organized the classical tracks of the IPC 2018.
SP6Q: Presenting a Paper
Eugene C. Freuder
The tutorial will provide advice on making an effective, enjoyable and memorable presentation of a scientific paper. The talk is an advertisement for the paper. You may have spent years working on your paper; you do not want to waste the few minutes you have in a conference presentation to interest your audience in what you have done. If your audience is checking their email because they are lost or bored, you are doing yourself and them a disservice. Presentation skills can be studied and practiced like any other. They can help prevent your paper from becoming WORN (Write Once Read Never).
Eugene Freuder
University College Cork
Eugene Freuder is a professor at the Insight Centre for Data Analytics, School of Computer Science & Information Technology, University College Cork, Cork, Ireland. He is an AAAS, AAAI, and EurAI Fellow, a Member of the Royal Irish Academy, an AAAI Councilor, and Presentation Chair of AAAI-19. He does not claim to be the world’s best speaker; however, he can share his own efforts to cope with a lack of natural speaking talent.
SP7Q: Planning and Scheduling Approaches for Urban Traffic Control
Scott Sanner, Mauro Vallati, and Stephen Smith
The current increase in urbanisation, coupled with the socio-economic motivation for increasing mobility, is pushing the transport infrastructure well beyond its capacity. Traditional urban traffic control techniques are struggling to cope with the dramatic rise of traffic and have limited ability to react. In response, more intelligent control mechanisms are required to better monitor and exploit the available infrastructure. Recently, the application of Automated Planning and Scheduling techniques to help in the management of road traffic has been investigated, and a number of approaches are now exploited or ready-to-be exploited in urban areas. The goal of the tutorial is to give an overview of the Planning & Scheduling techniques that have been designed to support Urban Traffic Control.
Scott Sanner
University of Toronto
Scott Sanner is an Assistant Professor in Industrial Engineering and Cross-appointed in Computer Science at the University of Toronto. Previously Scott was an Assistant Professor at Oregon State University and a Principal Researcher at NICTA. Scott earned a Ph.D. in Computer Science from the University of Toronto in 2008.
Mauro Vallati
University of Huddersfield
Mauro Vallati is a Senior Lecturer at the University of Huddersfield. His research focuses mainly on Automated Planning, Knowledge Engineering, and Abstract Argumentation. He is playing a significant role in SimplifAI, a company exploiting AI technologies for performing smart urban traffic control.
Stephen F. Smith
Carnegie Mellon University
Stephen F. Smith is a Research Professor of Robotics at Carnegie Mellon University, where his research focuses broadly on the theory and practice of next-generation technologies for automated planning and scheduling. He is also Founder/Chief Scientist of Rapid Flow Technologies; a company focused on smart transportation infrastructure for urban mobility.
SP8Q: The Road to Industry
John Kolen
Most doctoral AI programs focus on preparing students for academic careers. Given the relatively few academic openings compared to the number of graduates, the majority of freshly minted AI PhDs will find themselves in industry positions. This tutorial aims to ease this transition, providing information that will prepare graduates for this next phase of their lives. This tutorial is not a “how to write a resume” session but a collection of insights, motivations, and trivia that fall into the bucket labeled “I wish I knew that when I started.” The tutorial is organized into three parts: Finding the Job, Starting the Job, and Keeping the Job. Topics will include company types and the roles within them, the hiring process from the inside, acronyms you should know and what to do when you don’t, and life balance. The tutorial will finish with a question-and-answer panel session.
John Kolen
Electronic Arts
John Kolen’s interests include artificial intelligence, distributed systems, neural networks, and cognitive science. His career has threaded through academia, government, and industry. He is currently Head of Intelligent Systems at Electronic Arts, finding ways to leverage AI and data-driven solutions with their games. Before EA, he was deep in the bowels of web search at Google.
MA1: Adversarial Machine Learning
Bo Li, Dawn Song, and Yevgeniy Vorobeychik
Machine learning has seen a remarkable rate of adoption in recent years across a broad spectrum of industries and applications. Many applications of machine learning techniques are adversarial in nature insofar as the goal is to distinguish instances that are “bad” from those which are “good.” Indeed, adversarial use goes well beyond this simple classification example: forensic analysis of malware, which incorporates clustering, anomaly detection, and even vision systems in autonomous vehicles, could all potentially be subject to attacks. In response to these concerns, there is an emerging literature on adversarial machine learning, which spans both the analysis of vulnerabilities in machine learning algorithms and algorithmic techniques which yield more robust learning.
This tutorial will survey a broad array of these issues and techniques from both the cybersecurity and machine learning research areas. In particular, we consider the problems of adversarial classifier evasion, where the attacker changes behavior to escape being detected, and poisoning, where training data itself is corrupted. We discuss both the evasion and poisoning attacks, first on classifiers and then on other learning paradigms, and the associated defensive techniques. We then consider specialized techniques for both attacking and defending neural networks, particularly focusing on deep learning techniques and their vulnerabilities to adversarially crafted instances.
Bo Li
University of Illinois at Urbana–Champaign
Bo Li is an Assistant professor in the Department of Computer Science at the University of Illinois at Urbana–Champaign. Her research interest lies in adversarial deep learning, security, privacy, and game theory. She has developed and analyzed scalable, robust learning frameworks for learning algorithms in adversarial environments against evasion attacks. She has also analyzed adversarial behavior against learning algorithms in the physical world. She was a recipient of a Symantec Research Labs Graduate Fellowship. She obtained her Ph.D. degree from Vanderbilt University in 2016.
Dawn Song
UC, Berkeley
Dawn Song is a Professor in the Department of Electrical Engineering and Computer Science at UC Berkeley. Her research interest lies in deep learning and security. She has studied diverse security and privacy issues in computer systems and networks, including areas ranging from software security, networking security, database security, distributed systems security, and applied cryptography to the intersection of machine learning and security. She is the recipient of various awards, including the MacArthur Fellowship, the Guggenheim Fellowship, the NSF CAREER Award, the Alfred P. Sloan Research Fellowship, the MIT Technology Review TR-35 Award, the George Tallman Ladd Research Award, the Okawa Foundation Research Award, the Li Ka Shing Foundation Women in Science Distinguished Lecture Series Award, the Faculty Research Award from IBM, Google, and other major tech companies, and Best Paper Awards from top conferences. She obtained her Ph.D. degree from UC Berkeley. Prior to joining UC Berkeley as a faculty, she was an Assistant Professor at Carnegie Mellon University from 2002 to 2007.
Yevgeniy Vorobeychik
Washington University in St. Louis
Yevgeniy Vorobeychik is an Associate Professor at the Department of Computer Science and Engineering at Washington University in St. Louis. Previously, he was a Principal Research Scientist at Sandia National Laboratories. Between 2008 and 2010, he was a post-doctoral research associate at the University of Pennsylvania Computer and Information Science department. He received Ph.D. (2008) and M.S.E. (2004) degrees in Computer Science and Engineering from the University of Michigan and a B.S. degree in Computer Engineering from Northwestern University. His work focuses on game theoretic modeling of security and privacy, adversarial machine learning, algorithmic and behavioral game theory and incentive design, optimization, agent-based modeling, complex systems, network science, and epidemic control. Dr. Vorobeychik received an NSF CAREER award in 2017 and was invited to give an IJCAI-16 early career spotlight talk. He was nominated for the 2008 ACM Doctoral Dissertation Award and received an honorable mention for the 2008 IFAAMAS Distinguished Dissertation Award.
MA2: Deep Bayesian and Sequential Learning
Jen-Tzung Chien
The presentation of this tutorial is arranged into five parts. First, we share the current research on natural language processing, statistical modeling, and deep neural network and explain the key issues in deep Bayesian learning for discrete-valued observations and latent semantics. Modern natural language models are introduced to address how data analysis is performed, from language processing to semantic learning and memory networking. Secondly, we address a number of Bayesian models to infer hierarchical, thematic, and sparse topics from natural language. In the third part, a series of deep models, including deep unfolding, Bayesian recurrent neural network (RNN), sequence-to-sequence learning, convolutional neural network, generative adversarial network, and variational auto-encoder are introduced. The fourth part illustrates how deep Bayesian learning is developed to infer the recurrent and dilated neural networks for natural language understanding. In particular, the memory network, neural variational learning, and Markov RNN are introduced for practical tasks, e.g., speech recognition, reading comprehension, sentence generation, dialogue system, question answering, and machine translation. In the final part, we spotlight some future directions and challenges with big data, heterogeneous conditions and dynamic systems. In particular, deep learning, structural learning, temporal and spatial modeling, long-history representation, and stochastic learning are emphasized.
Jen-Tzung Chien
National Chiao Tung University, Taiwan
Jen-Tzung Chien is the Chair Professor at the National Chiao Tung University, Taiwan. He has served as the tutorial speaker for ICASSP, INTERSPEECH, and COLING. Dr. Chien has published extensively, including two books and ~200 peer-reviewed articles, many on Bayesian learning, sequential learning, deep learning, speech, and natural language processing.
MA3: Multi-Agent Pathfinding: Models, Solvers, and Systems
Roman Barták, Philipp Obermeier, Torsten Schaub, Tran Cao Son, Roni Stern
A fundamental task when designing physical multi-agent systems is pathfinding: how to move the agents from one location to another safely. While pathfinding for a single agent can be solved optimally in polynomial time, pathfinding for multiple agents is significantly more difficult. Nevertheless, recent years have shown tremendous progress in solving this problem in practice using a range of techniques following different algorithmic approaches.
In this tutorial, we will cover existing approaches for solving the multi-agent pathfinding problem (MAPF), discuss the pros and cons of each approach, and outline current challenges and opportunities in the field. We will also give a demo of systems for MAPF and robotic intra-logistics.
The target audience is anyone interested in planning for multiple agents. The attendees will walk away with information about the state-of-the-art in MAPF. The prerequisite knowledge for this tutorial is basic knowledge of heuristic search (best-first search, A*), SAT, and constraint satisfaction (all at the level of a graduate AI course).
Roman Barták
Charles University in Prague (Czech Republic)
Roman Barták works as a full professor and a researcher at Charles University in Prague (Czech Republic), where he leads Constraint Satisfaction and Optimization Research Group. His research work focuses on techniques of constraint satisfaction and modeling and their application to planning, scheduling, and other areas.
Philipp Obermeier
University of Potsdam
Philipp Obermeier is a doctoral student in the Knowledge-Based Systems group at the University of Potsdam, Germany. His scientific interests are centered around Answer Set Programming (ASP) and declarative languages for dynamic domains.
Torsten Schaub
University of Potsdam
Torsten Schaub is a professor at the University of Potsdam, Germany. His current research focus lies on Answer set programming and materializes at potassco.org, the home of the open-source project Potassco bundling software for Answer Set Programming.
Tran Cao Son
New Mexico State University
Tran Cao Son is a full Professor at New Mexico State University, USA. His main research interests are centered around theoretical issues and practical applications of knowledge representation and reasoning, such as aggregates in answer set programming, planning with preferences and incomplete information, multi-agent systems, and reasoning about actions and changes.
Roni Stern
Ben Gurion University
Roni Stern is a senior lecturer in the Department of Software and Information Systems Engineering at Ben Gurion University. His main research interests are heuristic search, automated diagnosis, and single and multi-agent automated planning. He is also interested in the applications of AI to health care.
MA4: Neural Vector Representations beyond Words: Sentence and Document Embeddings
Gerard de Melo
While word embeddings such as those produced by word2vec and GloVe are widely known as simple means of working with textual data, there has recently been substantial progress on improved methods that yield better embeddings. In particular, one may wish to induce neural vector representations not just of individual words but also of longer units of language, including 1) multi-word phrases, 2) entire sentences, or even 3) complete documents. Algorithms for these settings can draw on large corpora but may also exploit supervision from other kinds of data, such as document labels, lexical resources, or natural language inference datasets. Sentence embeddings are of particular interest because they may need to properly account for quite subtle distinctions between overall rather similar sentences. Moreover, new techniques have been developed to develop embeddings for multilingual and cross-lingual settings. This tutorial will thus provide an overview of recent state-of-the-art methods that go beyond word2vec and better model the semantics of longer units, such as sentences and documents, both monolingually and cross-lingually. The tutorial will start with a brief refresher on word2vec and how it relates to classic methods for distributional semantics, so no prior knowledge is required.
Gerard de Melo
Rutgers University
Gerard de Melo is an Assistant Professor at Rutgers University, heading a team of researchers working on NLP and AI. He has published over 100 papers, winning Best Paper/Demo awards at WWW, CIKM, ICGL, and the NAACL VSM Workshop, among others. For more information, please consult http://gerard.demelo.org.
MA5: Recent Advances in Scalable Retrieval of Personalized Recommendations
Dung Le, Hady Lauw
Top-K recommendation seeks to deliver a personalized recommendation list of K items to users. The dual objectives are accuracy in identifying the items a user is likely to prefer and efficiency in constructing the recommendation list in real-time. Naively scanning millions of items to identify the few most relevant ones may inhibit truly real-time retrieval performance. Towards improving the retrieval efficiency, we formulate it as an approximate K-nearest neighbour search aided by indexing schemes, e.g., locality-sensitive hashing, spatial trees, an inverted index. These speed up the retrieval process by discarding a large number of potentially irrelevant items when given a user vector. However, many recommendation algorithms rely on the inner product as a predictor, producing representations that may not align well with the structural properties of these indexing schemes, eventually resulting in a significant loss of accuracy post-indexing. In this tutorial, we first provide a theoretical justification and empirical demonstration of the potential issues arising from the incompatibility of the inner product search and indexing-based retrieval of recommendations. We then summarize different approaches that attempt to build fast and accurate retrieval systems for personalized recommendations. Finally, we conduct a hands-on session on using various indexing data structures for efficient and accurate recommendation retrieval.
Dung D. Le
Singapore Management University
Dung D. Le is a Ph.D. candidate in the Information Systems program at Singapore Management University (SMU). His research interests include recommender systems and information retrieval, with publications in major venues such as CIKM and SDM. In his candidature, he has been recognized with SMU Presidential Doctoral Fellowship and Ph.D. Student Life Award.
Hady W. Lauw
Singapore Management University
Hady W. Lauw is an Associate Professor at SMU School of Information Systems. He leads the Preferred. AI project, whose research spans artificial intelligence and machine learning, focusing on preferences and recommendations. His research is supported by a Singapore National Research Foundation Fellowship. He received his Ph.D. from Nanyang Technological University in 2008.
MP1: End-to-end Goal-oriented Question Answering Systems
Deepak Agarwal, Bee-Chung Chen, Qi He, Jaewon Yang, Liang Zhang
In this tutorial, we introduce goal-oriented conversational AI systems – guiding users to complete a specific task like finding a job via a conversational bot. We first introduce a variety of conversational AI systems based on knowledge graphs and intent classification proposed by pioneer researchers. The audience can easily comprehend what common technical components remain challenging and what unique engineering heuristics are useful. For the first time, the audience can learn in-depth not only the scientific methods that boost the precision of question understanding and answer retrieval/generation but also our practical experiences as well as engineering designs that enable an end-to-end system. Then, on top of three LinkedIn real scenarios, we share our hands-on experiences in the end-to-end process of building goal-oriented bots, including problem analysis from scratch, architecture design, training data collection, paraphrase generation, intent modeling, and dialogue management. Our goal is that, after this tutorial, the audience knows how to efficiently build a goal-oriented bot without getting stuck in unrealistic solutions. To attend this tutorial, you need to know basic machine learning knowledge like classification, basic deep learning knowledge like CNN/RNN, and basic natural language understanding knowledge like lexicon, and semantic parser.
Deepak Agarwal
Deepak Agarwal is VP of Artificial Intelligence at LinkedIn, heading the whole LinkedIn AI org. Fellow of the American Statistical Association, Member Board of Directors for SIGKDD, program chair of KDD in the past, and associate editor of two top-tier journals in Statistics. He has 20+ years of AI experience.
Bee-Chung Chen
Bee-Chung Chen is Principal Staff Engineer at LinkedIn with extensive industrial and research experience in recommender systems, machine learning algorithms, and big data processing. Some of his work is summarized in the book “Statistical Methods for Recommender Systems”. He currently leads the development of machine learning technology for the entire company.
Qi He
Qi He is the Director of Engineering at LinkedIn, heading the data standardization and knowledge graph team. Prior to LinkedIn, Research Staff Member at IBM Almaden, a postdoc at PSU, PhD from NTU. He is a steering committee member of CIKM, associate editor of TKDE, and the recipient of the 2008 SIGKDD Best Application Paper award.
Jaewon Yang
Jaewon Yang is a Staff Software Engineer at LinkedIn, where he leads projects on conversational AI. Jaewon obtained his Ph.D. degree at Stanford University in 2014. He received SIGKDD Doctoral Dissertation Award honorable mention in 2014.
Liang Zhang
Liang Zhang is the Head of Search Relevance, and Director of AI Engineering at LinkedIn, who has built many cutting-edge AI technologies in various LinkedIn products. Liang has a Ph.D. degree in Statistics from Duke University. He has published extensively in top-tier computer science conferences and also co-authored 20+ AI-related patents.
MP2: Graph Representation Learning
Dung Le, Hady Lauw
Machine learning on graphs is an important and ubiquitous task with applications ranging from drug design to friendship recommendation in social networks. The primary challenge in this domain is finding a way to represent or encode graph structure so that it can be easily exploited by machine learning models. Traditionally, machine learning approaches relied on user-defined heuristics to extract features encoding structural information about a graph (e.g., degree statistics or kernel functions). However, recent years have seen a surge in approaches that automatically learn to encode graph structure into low-dimensional embeddings, using techniques based on deep learning and nonlinear dimensionality reduction. In this tutorial, we provide a technical introduction to key advancements in this area of representation learning on graphs, including matrix factorization-based methods, random-walk-based algorithms, graph neural networks, and graph generation. In doing so, we develop a unified framework to describe these recent approaches, and we emphasize real-world applications involving large-scale social and biological networks.
The tutorial is targeted towards a general machine learning audience and will assume familiarity with common deep learning methods, such as LSTMS. However, we expect that experts in graph representation learning will also benefit from the tutorial’s synthesis of disparate techniques.
William Hamilton
Facebook AI Research
William Hamilton (wlh@fb.com) is a Visiting Researcher at Facebook AI Research, and he will be joining McGill University’s School of Computer Science as an Assistant Professor in January 2019. He completed his Ph.D. at Stanford University in 2018. His research focuses on graph representation learning and large-scale computational social science.
Jian Tang
Montreal Institute for Learning Algorithms
Jian Tang is currently an assistant professor at Montreal Institute for Learning Algorithms (Mila) and HEC Montreal. He finished his Ph.D. at Peking University in 2014, was a researcher at Microsoft Research between 2014-2016, and was a Postdoc fellow at the University of Michigan and Carnegie Mellon University between 2016-2017.
MP3: Imagination Science: Beyond Data Science
Sridhar Mahadevan
This tutorial introduces a new field of study called imagination science in AI and CS. In contrast to data science, which constructs statistical summaries of experience and answers the question “What is,” imagination science explores a much broader palette of questions, including interventional questions like “What if,” counterfactual queries, and “Why” explanations. A wide spectrum of challenging real-world problems, from digital marketing to AI for social good, involve imagination science. The tutorial will describe a number of converging lines of research in AI, which can be viewed as early attempts to build imagination machines, ranging from generative adversarial networks to layered causal architectures that combine observation, intervention, and counterfactual reasoning. The tutorial will discuss several novel research directions, including new ways of modeling sequential decision-making using imagination models, as well as emerging new ideas that attempt to integrate causal knowledge and imagination models, such as GANs. The tutorial will also cover mathematical formalisms that go beyond conventional approaches, such as gradient descent, for training imagination machines, and build on ideas from economics, psychology, and cognitive science. The tutorial will also explore how research on imagination will lead to new architectures for the next generation of “What if” high-fidelity simulation engines, which will replace today’s “What is” search engines and lead to new ways of building the next generation of the world-wide web.
Sridhar Mahadevan
Adobe Research
Sridhar Mahadevan directs the Data Science Lab at Adobe Research, San Jose, and is an Adjunct Professor of computer science at the College of Information and Computer Sciences at the University of Massachusetts, Amherst. He was elected a Fellow of AAAI in 2014 for significant contributions to machine learning.
MP4: Integrating Human Factors into AI for Fake News Prevention: Challenges and Opportunities
Amulya Yadav, Aiping Xiong
This tutorial aims to sketch the shape of an interdisciplinary approach involving artificial intelligence (AI) and cognitive psychology for fake news prevention on social media platforms. One objective is to illustrate how representations of human behavior can lead to realistic models/algorithms which address different aspects of the fake news problem, e.g., characterization, detection, and mitigation of fake news. Another is to emphasize the importance of understanding/characterizing the interactions between humans and these AI-aided systems.
The tutorial consists of three parts. In the two parts, we will present summaries of the latest work in the AI and cognitive psychology communities for fake news prevention, respectively. In the third part, we will propose an approach to integrating AI & cognitive psychology to alleviate the limitations of prior research. We will point out tangible research questions that would arise from this integration and propose possible solutions.
This tutorial is geared toward graduate students, AI researchers, and practitioners who are interested in fake news detection and prevention but do not have much background in human factors and want to learn about principles of human information processing and apply those principles to AI algorithms to detect and prevent fake news.
Amulya Yadav
Penn State University
Amulya Yadav is an Assistant Professor in the College of Information Sciences and Technology at Penn State University. His research interests include Artificial Intelligence, Multi-Agent Systems, Computational Game Theory, and Applied Machine Learning.
Aiping Xiong
Penn State University
Aiping Xiong is an Assistant Professor in the College of Information Sciences and Technology at Penn State University. Her research has focused on examining decision-making and human action selection within various cyber security and privacy contexts, including phishing, password generation, app selection, and autonomous driving.
MP5: Knowledge-based Sequential Decision-Making under Uncertainty
Shiqi Zhang, Mohan Sridharan
In this tutorial, we will focus on work at the intersection of declarative representations and probabilistic representations for reasoning and learning. There is significant prior work in probabilistic sequential decision-making (SDM) and in declarative methods for knowledge representation and reasoning (KRR). In this tutorial, we will highlight the complementary capabilities of these methods, summarize existing research that combines these capabilities, and identify some open problems in the design and use of such integrated systems in different application domains. We will focus on the interplay between goal-oriented SDM and declarative KRR and demonstrate how this interplay provides novel opportunities for addressing the open problems in the individual research areas. We will draw upon our own expertise in developing architectures that exploit these complementary capabilities for robots interacting and collaborating with each other and with humans. Our goal is to encourage many more researchers to explore the integration of probabilistic SDM and declarative KRR methods in different application domains. This tutorial will thus be of interest to researchers in these areas and in application domains such as robotics, computer vision, and natural language processing.
Shiqi Zhang
SUNY Binghamton
Shiqi Zhang is an Assistant Professor at the Department of Computer Science at the State University of New York (SUNY) at Binghamton. He was a Postdoctoral Fellow at the University of Texas at Austin (2014-2016) and received his Ph.D. in Computer Science (2013) from Texas Tech University.
Mohan Sridharan
University of Birmingham (UK)
Mohan Sridharan is a Senior Lecturer in the School of Computer Science at the University of Birmingham (UK). He received his Ph.D. from The University of Texas at Austin (USA). His research interests include knowledge representation and reasoning, machine learning, computational vision, and cognitive systems as applied to human-robot collaboration.
MP6Q: Human Identification at a Distance by Gait Analysis
Shiqi Yu, Yongzhen Huang, Yasushi Makihara, Daigo Muramatsu, Liang Wang, Yasushi Yagi, Tieniu Tang
Human identification at a distance is a very challenging task, which has long been a popular research topic in the field of computer vision. The gait sequences of different people can be very distinctive, which makes gait an important body characteristic that can be used for human identification. However, to the best of our knowledge, currently, there are very few tutorials concerning this important area. In the tutorial, we will introduce the history and the recent great progress in this area. They include the challenging problems in gait recognition, such as benchmarks, segmentation, feature representation, and recognition.
Shiqi Yu
Shenzhen University, China
Shiqi Yu received his B.E. degree in computer science and engineering from the Chu Kochen Honors College, Zhejiang University, in 2002 and his Ph.D. degree in pattern recognition and intelligent systems from the Institute of Automation, Chinese Academy of Sciences in 2007. He worked as an assistant professor and then as an associate professor at the Shenzhen Institutes of Advanced Technology, Chinese Academy of Science, from 2007 to 2010. Currently, he is an associate professor at the College of Computer Science and Software Engineering at Shenzhen University, China. He is a member of the council of the China Society of Image and Graphics and served as the program chair of the 12th Chinese Conference on Biometric Recognition 2017 and the organization chair of the IAPR/IEEE Winter School on Biometrics 2018 and 2019. He especially focuses on gait recognition since 2003.
Yongzhen Huang
National Laboratory of Pattern Recognition (NLPR)
Yongzhen Huang received a B.E. degree from Huazhong University of Science and Technology in 2006 and a Ph.D. degree from the Institute of Automation, Chinese Academy of Sciences (CASIA) in 2011. Then he joined the National Laboratory of Pattern Recognition (NLPR) as an Assistant Professor in July 2011 and became an Associated Professor in Nov. 2013. His research interests include computer vision, pattern recognition, and machine learning. He has published one book and more than 60 papers in international journals and conferences such as IEEE TPAMI, IJCV, IEEE TIP, IEEE TMSCB, IEEE TCSVT, IEEE TMM, CVPR, ICCV, NIPS, and AAAI. He has obtained several honors and awards, including the Excellent Doctoral Thesis of Chinese Association for Artificial Intelligence (2012), the Best Student Paper of Chinese Conference on Computer Vision (2015), the Champion of PASCAL VOC Challenges on object detection (2010 and 2011), the Champion of Internet Contest for Cloud and Mobile Computing on Human Segmentation (2013), and the Second Prize and the Prize of Highest Accuracy with Low Energy in LPIRC (Low-Power Image Recognition Challenge) (2015). Dr. Huang is now a Senior Member of IEEE. He has served as Associate Editor of Neurocomputing, the web chair of AVSS2012, the publicity chair of CCPR2012, the program committee member of 6 conferences, and the peer reviewer of over 20 journals and conferences.
Yasushi Makihara
Osaka University
Yasushi Makihara received B.S., M.S., and Ph.D. degrees in Engineering from Osaka University in 2001, 2002, and 2005, respectively. He is currently an Associate Professor of the Institute of Scientific and Industrial Research, Osaka University. His research interests are computer vision, pattern recognition, and image processing including gait recognition, pedestrian detection, morphing, and temporal super resolution. He is a member of IPSJ, IEICE, RSJ, and JSME. He has obtained several honors and awards, including the 2nd Int. Workshop on Biometrics and Forensics (IWBF 2014), IAPR Best Paper Award, the 9th IAPR Int. Conf. on Biometrics (ICB 2016), Honorable Mention Paper Award, the 28th British Machine Vision Conf. (BMVC 2017), Outstanding Reviewers, the 11th IEEE Int. Conf. on Automatic Face and Gesture Recognition (FG 2015), Outstanding Reviewers, and the 30th IEEE Conf. on Computer Vision and Pattern Recognition (CVPR 2017), Outstanding Reviewers. He has served as an associate editor of IPSJ Transactions on Computer Vision and Applications (CVA), a program co-chair of the 4th Asian Conf. on Pattern Recognition (ACPR 2017), and reviewers of CVPR, ICCV, ECCV, ACCV, ICPR, FG, etc.
Daigo Muramatsu
Osaka University
Daigo Muramatsu received the B.S.,丂M.E., and Ph.D. degrees in electrical, electronics, and computer engineering from Waseda University, Tokyo, Japan, in 1997, 1999, and 2006, respectively. He is currently an Associate Professor at The Institute of Scientific and Industrial Research at Osaka University. His current research interests include gait recognition, signature verification, and biometric fusion. He is a member of the IEEE, IEICE, and the IPSJ. He has obtained several honors and awards, including the 2nd Int. Workshop on Biometrics and Forensics (IWBF 2014), IAPR Best Paper Award, the 9th IAPR Int. Conf. on Biometrics (ICB 2016), Honorable Mention Paper Award, and the 11th Int. Workshop on Robust Computer Vision (IWRCV 2016), Best Poster Honorable Mention Award.
Liang Wang
Chinese Academy of Sciences
Liang Wang received both the B. Eng. and M. Eng. degrees in Electronic Engineering from the Department of Electronics Engineering and Information Science, Anhui University (AHU), China, in 1997 and 2000, respectively, and the Ph.D. degree in Pattern Recognition and Intelligent System from the National Laboratory of Pattern Recognition (NLPR), Institute of Automation, Chinese Academy of Sciences (CAS), China, in 2004. After graduation, he has worked as a Research Assistant at the Department of Computing, Imperial College London, United Kingdom, and at the Department of Electrical and Computer Systems Engineering, Monash University, Australia, and as a Research Fellow at the Department of Computer Science and Software Engineering, University of Melbourne, Australia, respectively. Before he returned back to China, he was a Lecturer with the Department of Computer Science, University of Bath, United Kingdom. Currently, he is a Professor of the Hundred Talents Program at the Institute of Automation, Chinese Academy of Sciences, P. R. China. His major research interests include machine learning, pattern recognition, computer vision, multimedia processing, and data mining. He has widely published in highly-ranked international journals such as IEEE TPAMI, IEEE TIP, IEEE TKDE, IEEE TCSVT, IEEE TSMC, CVIU, and PR, and leading international conferences such as CVPR, ICCV, and ICDM. He has obtained several honors and awards, such as the Special Prize of the Presidential Scholarship of the Chinese Academy of Sciences and the Research Commendation from the University of Melbourne in recognition of Excellent Research. He is currently a Senior Member of IEEE (Institute of Electrical and Electronics Engineers), as well as a member of the IEEE Computer Society, IEEE Communications Society, IEEE Signal Processing Society, and BMVA (British Machine Vision Association). He is serving with more than 20 major international journals and more than 40 major international conferences and workshops. He is an associate editor of IEEE Transactions on Systems, Man and Cybernetics-Part B, IEEE Transactions on Information and Forensic Security, International Journal of Image and Graphics (WorldSci), International Journal of Signal Processing (Elsevier), Neurocomputing (Elsevier), and International Journal of Cognitive Biometrics (Inderscience). He is a leading guest editor of several special issues such as PRL (Pattern Recognition Letters), IJPRAI (International Journal of Pattern Recognition and Artificial Intelligence), IEEE TIFS, and IEEE TSMC-B, as well as a co-editor of 6 edited books. He has also co-chaired one invited special session and eight international workshops. He was a co-PC chair of the 9th IEEE AVSS 2012.
Yasushi Yagi
Osaka university
Yasushi Yagi is the Executive Vice President of Osaka University. He received his Ph.D. degree from Osaka University in 1991. In 1985, he joined the Product Development Laboratory of Mitsubishi Electric Corporation, where he worked on robotics and inspections. He became a Research Associate in 1990, a Lecturer in 1993, an Associate Professor in 1996, and a Professor in 2003 at Osaka University. He was also Director of the Institute of Scientific and Industrial Research at Osaka University from 2012 to 2015. International conferences for which he has served as Chair include FG1998 (Financial Chair), OMINVIS2003 (Organizing chair), ROBIO2006 (Program co-chair), ACCV2007 (Program chair), PSVIT2009 (Financial chair), ICRA2009 (Technical Visit Chair), ACCV2009 (General chair), ACPR2011 (Program co-chair) and ACPR2013 (General chair). He has also served as the Editor of the IEEE ICRA Conference Editorial Board (2007–2011). He is the Editorial member of IJCV and the Editor-in-Chief of IPSJ Transactions on Computer Vision & Applications. He was awarded the ACM VRST2003 Honorable Mention Award, IEEE ROBIO2006 Finalist of T.J. Tan Best Paper in Robotics, IEEE ICRA2008 Finalist for Best Vision Paper, MIRU2008 Nagao Award, and PSIVT2010 Best Paper Award. His research interests are computer vision, medical engineering, and robotics. He is a fellow of IPSJ and a member of IEICE, RSJ, and IEEE.
Tieniu Tan
National Laboratory of Pattern Recognition (NLPR)
Tieniu Tan received his BSc degree in electronic engineering from Xi’an Jiaotong University, China, in 1984 and his MSc and Ph.D. degrees in electronic engineering from Imperial College London, U.K., in 1986 and 1989, respectively. In October 1989, he joined the Computational Vision Group at the Department of Computer Science, The University of Reading, Reading, U.K., where he worked as a Research Fellow, Senior Research Fellow, and Lecturer. In January 1998, he returned to China to join the National Laboratory of Pattern Recognition (NLPR), Institute of Automation of the Chinese Academy of Sciences (CAS), Beijing, China, where he is currently a Professor and the director of the Center for Research on Intelligent Perception and Computing (CRIPAC) and was former director (1998-2013) of the NLPR and Director General of the Institute (2000-2007). He is currently also Deputy Director of the Liaison Office of the Central People’s Government in the Hong Kong S.A.R. He has published 14 edited books or monographs and more than 600 research papers in refereed international journals, and conferences in the areas of image processing, computer vision, and pattern recognition. His current research interests include biometrics, image and video understanding, and information content security.