MUMBAI, India, Sept. 14 -- Intellectual Property India has published a patent application (202621089366 A) filed by Ramdeobaba University, Nagpur; Shri Ramdeobaba College Of Engineering And Management, Nagpur; Bani, Supriya Gupta; Rathi, Khushi; Khanke, Uttkarsh; and Vyas, Devesh on July 22, 2026, for A System For Self-Adaptive Stock Market Prediction.
Inventors include Bani, Supriya Gupta; Rathi, Khushi; Khanke, Uttkarsh; and Vyas, Devesh.
The application for the patent was published on September 11, 2026, under issue no. 37/2026.
Abstract: The present invention relates to a system (100) for self-adaptive stock market prediction. The system includes a data input and monitoring module (110), a hybrid prediction module (120), an error monitoring module (130), a concept drift detection module (140), an adaptive retraining module (150), a dynamic selection module (160), a model reconfiguration engine (170), and an output module (180). The data input and monitoring module (110) receives real-time stock market data or historical stock market data and maintains a sliding window of recent stock market data. The hybrid prediction module (120) generates stock market predictions using a SARIMA model (122) and a recurrent neural network model (124). The error monitoring module (130) computes prediction error values and rolling error statistics. Upon detection of concept drift, the models are retrained, a best-performing model configuration is selected and reconfigured, and updated predictions, model performance status, and concept drift alerts are presented.
Disclaimer: Curated by HT Syndication.
