Unsupervised Feature Extraction Using Singular Value Decomposition

Abstract

Though modern data often provides a massive amount of information, much of the insight might be redundant or useless (noise). Thus, it is significant to recognize the most informative features of data. This will help the analysis of the data by removing the consequences of high dimensionality, in addition of obtaining other advantages of lower dimensional data such as lower computational cost and a less complex model. Modern data has high dimension, sparsity and correlation besides its characteristics of being unstructured, distorted, corrupt, deformed, and massive. Feature extraction has always been a major toll in machine learning applications. Due to these extraordinary features of modern data, feature extraction and feature reduction models and techniques have even more significance in analyzing and understanding the data.