A Review of Various Dimension Reduction Methods

Data analysis and prediction become an indispensable part of many fields. However, the data with high dimensionality may cause problems, such as memory waste, or data loss while processing. In this case, dimension reduction is necessary for data processing. In this paper, a number of different dimension reduction methods have been discussed, including the theoretical background and simple examples of how they perform.

I. Introduction

In this era, data processing is a commonly used technology, and it is essential in a variety of fields. However, the raw data that has been generated may have a great number of features, the different features of data are called the dimension of data. It is challenging to analyze the data with high dimensions, and it could take a massive amount of time and computational cost. Thus, reducing the dimensionality of data while keeping the important features become crucial. In this paper, a variety of dimension reduction method has been reviewed and tested.