An Optimization of Feature Selection for Classification using Bat Algorithm

Data mining is the action of searching large existing database in order to get new and best information. It plays a major vital role now-a-days all sorts fields like Medical, Engineering, Banking, Education Fraud detection. In this paper Feature selection which part performed do classification. The feature context deep learning how it related engineering. preprocessing technique selects appropriate features from data set accurate result outcome for Natureinspired Optimization algorithms Ant colony, Firefly, Cuckoo Search Harmony showed better performance by giving accuracy rate with less number selected also fine f-Measure value noted. These are used perform classification that accurately predicts target class each case set. We propose optimized using Meta Heuristic algorithms. applied recent advanced algorithm named Bat on UCI datasets comparatively equal results firefly but selected. work implemented JAVA Medical dataset (UCI) has been used. were chosen due nominal features. attributes, instances classes varies represent different combinations. Classification done J48 classifier WEKA tool. demonstrate comparative presently thoroughly.