Analysis of various approaches for stock market prediction

Abstract

Stock Market Prediction (SMP) is highly complex in nature and its research has been of utmost importance in recent years. People involved in the Stock Market do not invest randomly. They invest based on some kind of prediction. Employing traditional methods like technical and fundamental analysis may not ensure the reliability of the prediction. Although one can never be sure of the rise and fall of the Market, predicting it to a great extent is very much possible using the modern techniques of Machine Learning (ML), Data Mining and Deep Learning. In this paper, we survey various approaches including Support Vector Machine (SVM), Random Forests (RF), Naïve Bayes (NB), Regression and some fusion models. These modern methods which involve ML, Data Mining and Deep Learning have proven to give more reliable results than the traditional methods of Stock Prediction and have a high possibility of advancement in the future.

Subject Classification:

• 68N01

Keywords:

• Stock Prediction
• Regression
• Support Vector Machine
• Random Forests
• Naïve Bayes
• Data Mining
• Deep Learning
• Machine Learning