Intellectual Property India Publishes Patent Application for 'Lstm-Based Stock Volatility Prediction System For Emerging Markets' Filed by Vellore Institute Of Technology

MUMBAI, India, Jan. 2 -- Intellectual Property India has published a patent application (202541122972 A) filed by Vellore Institute Of Technology, Vellore, Tamil Nadu, on Dec. 5, 2025, for 'lstm-based stock volatility prediction system for emerging markets.'

Inventor(s) include Dr. Narayana Moorthi. M; Aditya Santosh Pande; and Gauri Sopan Dhamale.

The application for the patent was published on Jan. 2, under issue no. 01/2026.

According to the abstract released by the Intellectual Property India: "The present disclosure provides a computer-implemented method (100) for predicting stock price volatility in emerging markets. The method (100) includes collecting (102) historical stock data from emerging markets, training (110) a Long Short-Term Memory neural network model using the historical stock data, and generating (112) stock price volatility predictions using the trained model. The method (100) may further include preprocessing (106) the collected data by cleaning inconsistencies and handling missing values prior to training (110), and engineering (108) technical indicators from the historical data. The Long Short-Term Memory neural network model includes multiple layers with dropout regularization and attention mechanisms for focusing on relevant time periods. The generating (112) produces volatility forecasts for specified time horizons and calculates confidence intervals using parametric and non-parametric methods. The method (100) may also include displaying (114) the predictions through a web-based interface with interactive visualizations."

Disclaimer: Curated by HT Syndication.