I. INTRODUCTION
Dimensionality reduction aims to maintain the majority of the "intrinsic information" included in high-dimensional data samples by embedding them in low-dimensional spaces (e.g., Roweis and Saul, 2000; Tenenbaum et al., 2000; Hinton and Salakhutdinov, 2006). [7]. Data Reduction is a crucial step of pre-processing of the Dataset. Predictive models with redundant and irrelevant data perform poorly; this data should be deleted from training datasets to improve model performance [3]. In order to reduce noisy data and avoid overfitting, Data Reduction aids in the selection of the appropriate outcomes.