A Comparative Study of Regression and Evolution-Based Stock Selection Models for Investor Sentiment

摘要: Stock selection has long been recognized as an important task in finance. Researchers and practitioners this area often use regression models to tackle problem due their simplicity effectiveness. Recent advances machine learning (ML) are leading significant opportunities solve these problems more effectively. In paper, we present a comparative study between the traditional regression-based evolution-based using investor sentiment indicators for stock selection. models, Genetic Algorithms (GA) used optimization of model parameters feature input variables simultaneously. We will show that our proposed GA-based method significantly outperforms well benchmark. thus expect methodology advance research behavioral