MUMBAI, India, May 29 -- Intellectual Property India has published a patent application (202541053117 A) filed by Bharath Gaddam, Livingston, U.S.A., on May 31, 2025, for 'system and method for causal time series forecasting using pattern-oriented embedding models.'
Inventor(s) include Bharath Gaddam.
The application for the patent was published on May 29, under issue no. 22/2026.
According to the abstract released by the Intellectual Property India: "A system (100) for time series forecasting comprises of a server (102) comprising at least one processor (104) and a memory (106). The processor (104) comprises of a data pre processing module (108), a temporal processing module (110), a causal discovery module (112) and a hierarchical integration module (114) and a machine learning module (116). A method (200) for time series forecasting, comprising steps of: receiving multivariate time series data from multiple domains by a server (102), processing the data by a data preprocessor module (108), performing causal discovery by a causal discovery module (112), performing hierarchical integration by a hierarchical integration module (114), training a machine learning model by a machine learning module (116) and generating long-horizon forecasts by the machine learning module (116). The invention provides fundamental causal mechanisms, enabling superior forecasting accuracy, adaptability, and interpretability in complex, real-world time series environments."
Disclaimer: Curated by HT Syndication.
