Abstract: The classification and categorization is major issue in multi-features based data. As the classification and categorization of multi-features data are used multi-label classifications. The multi-label classification technique used the similarity based features selection process. In this paper, we proposed the feature optimization based multi-label classification. For the optimization of features used teacher learning based optimization technique. The teacher learning based optimization technique is basically based on dynamic iteration population algorithm and it reduces the unwanted and distinct feature of data for the process of classification. The modified algorithm implemented in mat lab software and used some reputed dataset for the evaluation of performance.

Keywords: Data Mining, MLC, Feature Optimization, TLBO