MUMBAI, India, Aug. 22 -- Intellectual Property India has published a patent application (202421011073 A) filed by G. H. Raisoni College Of Engineering; and G. H. R. Labs And Research Centre, Nagpur, Maharashtra, on Feb. 16, 2024, for 'a drought forecasting system.'
Inventor(s) include Gaurihar, Mahima M; Paonikar, Kaustubh R; Dongre, Snehalata; Khobragade, Prashant; Saraf, Pranay; and Agrawal, Rahul.
The application for the patent was published on Aug. 22, under issue no. 34/2025.
According to the abstract released by the Intellectual Property India: "The present invention is related to a drought forecasting system. This invention presents a groundbreaking Meta-Learning-Based Drought Forecasting System, combining Ridge Regression, Support Vector Machines, LSTM, and ARIMA as base learners (102), consolidated by a Random Forest (104) meta learner (105). An adaptive weighting mechanism dynamically assigns weights based on model performance, enhancing predictive accuracy. Validated using historical data, the system employs comprehensive performance evaluation metrics, including MSE, MAE, and R2. The hierarchical architecture, seamlessly implemented in a Jupyter notebook, ensures modularity and efficiency. The innovative meta-learning application in environmental science allows the system to adapt to new data patterns and machine learning algorithms, providing a scalable and adaptable solution. Results demonstrate superior performance, making this system a significant contribution to precise and actionable drought forecasting for researchers, policymakers, and farmers."
Disclaimer: Curated by HT Syndication.
