Real-Time Stock Trend Analysis and Investment Support Tool Using Deep Learning

Kalyani Sanap, Neha A. Rathod, Sakshi Alone, Achal Battulwar, Vrukshada Ukinkar, Manisha More

Abstract

Financial markets' volatility, driven by economic, political, and psychological factors, complicates stock price forecasting and investment decisions. This paper proposes a real-time stock trend analysis and investment support system using a hybrid deep learning ensemble of GRU, LSTM, and Transformer models with attention mechanisms for accurate price predictions from live API data (e.g., Yahoo Finance). It integrates K-means clustering for risk-based asset categorization (stocks, mutual funds, gold, bonds), decision trees for personalized buy/hold/sell recommendations, and portfolio optimization via MPT or reinforcement learning. An interactive dashboard visualizes trends, predictions, and insights, empowering users with data-driven, diversified strategies to enhance returns and manage risks.

Source: semanticscholar · PDF

Read the AI summary and key takeaways for traders on WOBR Quant Research.