MUMBAI, India, Jan. 23 -- Intellectual Property India has published a patent application (202521124738 A) filed by Gopal Dadarao Upadhye; Shital P. Dongre; Kriya Jain; Vallabh Kathar; Sakshi Lohote; and Vibhor Kumbhare, Latur, Maharashtra, on Dec. 10, 2025, for 'ml powered portfolio optimization system.'
Inventor(s) include Gopal Dadarao Upadhye; Shital P. Dongre; Kriya Jain; Vallabh Kathar; Vibhor Kumbhare; and Sakshi Lohote.
The application for the patent was published on Jan. 23, under issue no. 04/2026.
According to the abstract released by the Intellectual Property India: "Portfolio optimization in today's highly volatile markets needs models that are both flexible and strong. However, most existing systems depend on standard LSTM, Random Forest, or traditional MPT frameworks without any innovative algorithms or crisis-aware assessments. This paper presents IntelliPort, a hybrid optimization framework that combines a Regime-Aware Attention-LSTM (RA-LSTM) for better forecasting, a Hybrid Predictive-Risk Engine (H-PRE) that merges RA-LSTM, Random Forest regime detection, and Monte Carlo tail-risk simulation, along with a Drift-Aware Adaptive Rebalancing Algorithm (DARA) supported by a Behavior-Adjusted Rebalancing Index (BARI). Unlike earlier models, IntelliPort++ includes transaction-cost-aware CVaR constraints and tests performance across different market data sets (NSE, NASDAQ, BTC/ETH, gold) with out-of-sample evaluations during crisis periods. Compared to five leading models, this system shows statistically significant improvements in Sharpe ratio, drawdown reduction, and predictive accuracy. The results demonstrate that IntelliPort++ is a new, scalable, and IEEE-aligned contribution to real-time, machine-learning-driven portfolio optimization."
Disclaimer: Curated by HT Syndication.
