Feature Selection in Stock Market Prediction: A Comprehensive Review

Abstract

The current world is referred to as the "data world," as, according to Google, 328.77 million terabytes of data are generated every day and are            continually increasing. One of the cause contributing to the data growth is the stock market. Thus, it is now necessary to reduce data by removing unnecessary data and extracting only the data that is important. The feature selection               procedure is crucial for focusing on key data and reducing the dimensionality of the data. As per my knowledge there aren't many published articles that review the feature extraction and selection techniques utilized in stock market prediction at this time. The same motive will be covered in this paper, where we will analyze feature extraction and selection techniques utilized in the stock market. It              includes embedded, filter, wrapper, supervised, unsupervised, semi-supervised, and hybrid methods.