Stock market movement prediction using machine learning techniques and graph-based approaches

Abstract

Machine learning techniques are preferred now than the statistical methods for stock movement prediction due to their efficiency and effectiveness. Stock market movement prediction is impacted significantly by choice of input features and prediction algorithms. We focus on a specific event of ex-dividend day and use event-specific input features of cum-dividend period for predicting price movement on the ex-dividend day. Performance improves significantly when these event-specific optimum input features are used along with machine learning models. The relative order or ranking of stocks is more important than the price or return of a single stock for better investment decisions. Stock ranking performance can be improved by incorporating the stock relationship information in the prediction task. We employ a graph-based approach for stock ranking prediction and use the stock relationship information as the input of the machine learning model. Investing in the top-k stocks is more profitable than the others. Thus, the performance measure for stock ranking prediction should be top-weighted and bounded for any value of k. Existing evaluation measures lack these properties, and we propose normalized rank biased overlap for top-k (NRBO@k) stocks for stock ranking prediction. Moreover, we show that the list-wise loss function can significantly improve the stock ranking performance in a graph-based approach. We find that node embedding techniques such as Node2Vec can significantly reduce graph-based approaches’ training time for the stock ranking prediction. In our survey study, we discuss the existing graph-based works from five perspectives: i) stock market graph formulation, ii) stock market graph filtering, iii) stock market graph clustering, iv) stock movement prediction, and v) portfolio optimization. This survey contains a concise description of major techniques and algorithms relevant to graph-based approaches for the stock market.