This is my undergradudate thesis project codebase.
It is submitted to department of computer science and engineering, Bangladesh University of Engineering and Technology.
This thesis introduces a rule based framework for reality mining specially for detecting friendship between a pair, devises an evolutionary algorithm for learning generalized rules, evaluates the performance of this evolutionary approach on for the MIT's Reality Mining Dataset and compare the results with other techniques used so far for this Dataset. Also we use some UCI dataset to compare our approach with renowned algorithm. Portion of the paper is dedicated to the description of some previous approaches that are widely used in literature. It also presents some proposal of future work regarding the work done.
Evolutionary Algorithms (EA) are search algorithms based on the metaphor of "Natural Selection". During the last few decades, evolutionary algorithms have been applied successfully to solve particularly optimization problems from numerous and diverse application areas. Typically these algorithms maintain a population of individual solutions, each of which has a fitness attached to it, which in some way reflects the quality of the solution. The search proceeds via the iterative generation, evaluation and possible incorporation of new individuals based on the current population, using a number of parameterized genetic operators.
2012
[1]: Tarequl Islam Sifat [2]: Md. Ibrahim Rashid [3]: Muhammad Ali Nayeem
[1]: Dr. Md. Monirul Islam