Training a Stock Suggester Using Genetic Algorithm

Abstract

Training a stock suggester utilizing neural networks (CNN, RNN, LSTM) is challenging and has constraints. The model’s constraints include the need for computing power and a large dataset to train it. Neural network models range inaccuracy from bad to good, depending on the implementation. Is there a mechanism for creating trading bots with a similar level of accuracy? The chance to profit by investing in the stock market is difficult to achieve since it is constantly influenced by economic, political, and social issues. We use a different strategy to train the data set by employing genetic algorithms and evolutionary learning. This genetic algorithm will be at the root of a profitable stock suggester for the Indian stock markets (BSE, NSE).