A hybrid of fish swarm algorithm and shuffled frog leaping algorithm for attribute reduction
Zhiying Lu, Chenyi Wang, Jianlin Guo · 2018
Since data increases with time and space, attribute reduction (AR) becomes an important task in data mining application. In order to acquire better reduction results, a hybrid algorithm is presented by combing fish swarm algorithm and shuffled frog leaping algorithm (FSA-SFLA). It gives full play to the ability of FSA and SFLA. First, FSA is used for global optimization to find the region of optimal solution quickly. Then SFLA is used for local optimization, which has good local searching ability. The attribute reduction algorithm based on FSA-SFLA takes full advantages of these two swarm intelligent algorithms, which have faster convergence and higher efficiency. Some well-known datasets from UCI are used to verify the above-mentioned algorithm. Experiments demonstrate that the proposed method can effectively reduce attribute dimensions and keep, even improve classification ability.