DNA sequence based data classification technique

DNA sequence is a well established means for species classification and it motivated us to use DNA sequence for the purpose of data mining and knowledge discovery. In this work, we have considered different entities of DNA signature for the construction of an efficient classification algorithm. The proposed DNA sequence based data classification algorithm has been tested over three datasets namely: iris, wine and new-thyroid and its performance is observed satisfactory. Further, the performance of proposed algorithm has been compared with different variants of differential evolution (DE) and we observed the superiority of proposed method. These DE variants are used for optimizing the positioning of prototypes in nearest neighbour classification. We used the 1- NN rule as a baseline limit of performance.