Stock Trend Prediction

Abstract

Forecasting of stock market has been one of the most interesting researches for many scholars; the basic market analysis is an important component of it. may be connected to historical stock market data where a set of features can be generated. It will be imperative to pick features. it also provides the most relevant information on the aspect that is being discussed. This study investigates to use coefficient of variation (CV) - the selection of features for stock market prediction. Coefficient of variation (CV), a statistical technique that is commonly used to obtain is a widely used statistical method. variability among data distributions. CV is calculated for every feature and we are adding a k-means algo-rithm, median range and top-M, to select the feature set with a specific range of values. attributes that are the most important characteristics of the biggest cluster, and which have a defined range, and with the highest CV. values, respectively. We use the models selected features as the input and models like backpropagation neural network as the output. LSTM, GRU, CNN, BPNN are the most popular deep learning techniques for forecasting stock prices and trends.