The number of elderly people is constantly increasing in the developed countries. Elderly tend to lead an isolated life away from their offspring; however, they may fear being unable to obtain help if they are injured or ill. During the last decades, this fear has generated research attempts to find assistive technologies for making living of elderly people at homes easier and independent, as is the aim of this research work. Research study proposes a generalized approach to an intelligent and ubiquitous care system to recognize a few of the most common and important health problems of the elderly, which can be detected by analyzing their movement. In the event that the system was to recognize a health problem, it would automatically notify a physician with an included explanation of the automatic diagnosis. It is two-step approach; in the first step it classifies person's activities into five activities: fall, unconscious fall, walking, standing/sitting, lying down/lying. In the second step, it classifies walking patterns into five different health states; one healthy and four unhealthy: hemiplegia (usually the result of stroke), Parkinson’s disease, leg pain and back pain. Moreover, since elderly having these health problems are less stable and more prone to falls, recognizing them leads not only to detection but indirectly also to prevention of falls of elderly people. In the initial approach movement of the user is captured with the motion capture system, which consists of the tags attached to the body, whose coordinates are acquired by the sensors situated in the apartment. In the current approach wearable inertial sensors are used, allowing monitoring inside or outside of the buildings. Output time-series of coordinates are modeled with the proposed data mining approach to recognize the specific health problem.