A Survey on Stock Price Prediction using Machine Learning Techniques

In Today's competitive environment, every Industry wants to grow fast to become a leader in their business. There is a need for the regular growth of business in the form of expansion. For Expansion, they required more capital. There are primarily 3 ways to raise Capital. These ways are an Initial Public Offering (IPO), Angel investors, and Business loans. Generally, when a company grows beyond a certain size, it is difficult for individual investors to continue to run the operations within their own capacity of capital. Companies need constant capital. An IPO method for raising capital often raises a company's profile and enhances its credibility with suppliers, Creators, and customers. Raising capital is the primary motivation for an IPO, however, it can also provide an opportunity for a partial sell-down by founders/early-stage investors. After IPO, the shares of a company are available in the form of stocks in an open platform. Including stocks in a public investment portfolio can prove to be beneficial. By investing in stocks of various companies, individuals can generate savings and shield their money against inflation and taxes. However, in order to optimize returns on investment, it is important to have the ability to predict stock prices. All the surveyed techniques are based on the concepts of Machine Learning. These techniques are compared by highlighting their strengths and weaknesses. In the survey of existing works, it is found that combining LSTM with another model can be the most effective technique for stock price prediction. The future scope of improvements in the surveyed research is also suggested in the paper. Researchers in this field can take advantage of these suggested improvements for further research in the field to enhance the performance of techniques used for stock price prediction.