EPSO-GHSOM Stock Selecting and Trading Strategy on Big Data

Abstract

Extracting trading information from the stock market to construct accurate forecasting models that filter signals and noise is a challenge. This research employs big data analytics to construct a computation platform for stock selection and trading strategies. It adopts elite particle swarm optimization (EPSO) to elucidate optimal trading opportunities and combines growing hierarchical self-organizing map (GHSOM) and EPSO in its stock selection strategy. EPSO–GHSOM distinguishes companies’ operating profitability, identifies price signals, and sets decision rules for buying and selling.