A New Supervised t-SNE with Dissimilarity Measure for Effective Data Visualization and Classification

In this paper, a new version of supervised t-SNE algorithm is proposed which introduces using a dissimilarity measure elated with class information. The proposed S-tSNE can be applied in any high dimensional dataset for visualisation or as a feature extraction for classification problems. In this study the S-tSNE is applied to three datasets MNIST, Chest x-ray and SEER Breast Cancer. The two-dimensional data generated by the S-tSNE showed better visualization and an improvement in terms of classification accuracy in comparison to the original t-SNE method. The results from k-NN classification models which used the lower dimension space generated by the new S-tSNE methods showed more than 20% accuracy improvement in all the three datasets compared with t-SNE method. In addition, the classification accuracy using the S-tSNE for feature extraction was even higher than classification accuracy obtained from the original high-dimensional data.