Skip to Main Content
In this paper we propose FM-PGA, a MapReduce-based hybrid of FM-test and Parallel Genetic Algorithm (PGA), to analyze gene micro array data on cloud computing platform. We investigate the performance of FM-PGA on real-world micro array data and compare it with FM-GA and MapReduce PGA. The experimental results confirm that the genes selected by FM-PGA achieve comparable classification accuracy while the time to achieve that accuracy is significantly reduced.
Date of Conference: 7-9 Nov. 2011