Machine learning techniques and methodology analysis for stock market price prediction

Abstract

Prediction of market movements is a very emerging topic in research nowadays. The one handsome efficient hypothesis says that accurate stock prediction is impossible but, in another hand, if appropriate techniques and algorithms can be used then it can be predicted with high accuracy. Moreover, predicting profitable and accurate value is a challenging task for investors because price prediction is volatile in nature. It could be affected by various factors like the global economy, politics, disasters, etc. A lot of literature survey on technical analysis is available which identify the movement patterns of the market. Various hybrid machine learning approaches in combination have been used to forecast the beginning values of the market trends and also for the future long-term values. This paper gives a comparative assessment of 40 research papers that recommend strategies such as calculating methodologies, stock prediction algorithms, Datasets, results, and more which leads to improved accuracy and lowers the error percentage of stock market prediction.