Hybrid Forecasting and Optimization: ANN-GA Approach for Portfolio Optimization with Sentiment Analysis

  • Unique Paper ID: 181715
  • Volume: 12
  • Issue: 1
  • PageNo: 5034-5049
  • Abstract:
  • In recent years, individual participation in financial markets has gathered pace, encouraged by increased accessibility and a speeding focus on wealth management. With investors reacting to better asset allocation strategies for maximizing return while reducing risk, complex computational techniques have taken center stage. This research proposes a hybrid portfolio optimization model incorporating Artificial Neural Networks (ANN) for predicting stock return and risk, the Non-dominated Sorting Genetic Algorithm III (NSGA-III) for multi-objective optimization, and a FinBERT-based news sentiment analysis module for news-based risk adjustment. The model integrates dynamic market sentiment in a real-time manner using scraping company-specific news articles and examining sentiment scores for enhanced stock selection choices. Furthermore, true financial parameters are applied using real-world financial information obtained from the Yahoo Finance API (yfinance), and its perfor- mance is evaluated using comprehensive financial measures like Annualized Return, Annualized Volatility, Sharpe Ratio, Sortino Ratio, and Maximum Drawdown. NSGA-II and MOEA/D are compared apples to apples in order to provide optimization efficacy. Findings indicate that the method proves to be effective in constructing robust portfolios for different holding horizons, offering an extendable AI-driven sentiment-aware methodology to portfolio optimization.

Cite This Article

  • ISSN: 2349-6002
  • Volume: 12
  • Issue: 1
  • PageNo: 5034-5049

Hybrid Forecasting and Optimization: ANN-GA Approach for Portfolio Optimization with Sentiment Analysis

Related Articles