Affiliation:
1. Sarojini Naidu College for Women, India
Abstract
The prediction of stock prices is crucial for investment and risk management. Accurate forecasts give investors a competitive edge, enabling better trading decisions and optimized portfolio management. Despite many predictive methodologies, there is a lack of emphasis on analyzing the inherent complexity of stock price prediction itself. This chapter addresses this gap by quantifying the complexity of stock price prediction using regression complexity measures. Specifically, it focuses on the weekly stock price movements within the NIFTY 50 universe of India, utilizing ten years of historical weekly stock prices. The authors applied regression complexity measures to evaluate the complexity of predicting weekly stock prices and designed two scores to combine these measures. The findings show that these measures effectively differentiate between stocks with varying levels of prediction difficulty, enabling investors to make more informed decisions. Ranking stocks based on these measures helps investors select stocks with lower prediction complexity enhancing robust predictive models.
Reference26 articles.
1. The comparison of methods artificial neural network with linear regression using specific variables for prediction stock Price in Tehran stock exchange.;R. G.Ahangar;Int. J. of Comp Sc. and Informat Sec.,2010
2. Investigating the Performance of Data Complexity & Instance Hardness Measures as A Meta-Feature in Overlapping Classes Problem
3. AppelG. (2005). Technical analysis: power tools for active investors. FT Press.
4. The Dow Theory: William Peter Hamilton's Track Record Reconsidered
5. Hybrid deep learning diagonal recurrent neural network controller for nonlinear systems