Showing papers by "Hossam Faris published in 2012"

F ermentation Process Modeling Using Takagi-Sugeno Fuzzy Model

Abstract: Fermentation process is vital and important in many biotechnological applications. However modeling the fermentation process is considered a challenging and complex problem. The complexity of the problem is driven by the need of efficient, accurate, not expensive, and reliable predictive models. In this paper, we apply a Takagi-Sugeno Fuzzy Logic technique for modeling the lipase activity production based on nutritional and physico- chemical factors to estimate the lipase activities. The accuracy of the developed fuzzy model is measured, validated and compared with both the multiple regression and artificial neural network models. The fuzzy model successfully showed competitive and promising modeling results.

An Automatic Course Scheduling Approach Using Instructors' Preferences

Abstract: University Courses Timetabling problem has been extensively researched in the last decade. Therefore, numerous approaches were proposed to solve UCT problem. This paper proposes a new approach to process a sequence of meetings between instructors, rooms, and students in predefined periods of time with satisfying a set of constraints divided in variety of types. In addition, this paper proposes new representation for courses timetabling and conflict-free for each time slot by mining instructor preferences from previous schedules to avoid undesirable times for instructors. Experiments on different real data showed the approach achieved increased satisfaction degree for each instructor and gives feasible schedule with satisfying all hard constraints in construction operation. The generated schedules have high satisfaction degrees comparing with schedules created manually. The research conducts experiments on collected data gathered from the computer science department and other related departments in Jordan University of Science and Technology- Jordan.