MUMBAI, India, May 1 -- Intellectual Property India has published a patent application (202621028737 A) filed by Dr. Ravindra B. Gadhiya; Dr. Rakshitkumar S. Upadhyay; Dr Darshana B. Likhada; Dr. Rahulsinh B. Chauhan; Chaudhari Manisha Maheshchandra; Gauravkumar P Rathod; and Dehlvi Mohammadnaved J, Gandhinagar, Gujarat, on March 11, for 'a system and a method for automated feature extraction and learning optimization.'
Inventor(s) include Dr. Ravindra B. Gadhiya; Dr. Rakshitkumar S. Upadhyay; Dr Darshana B. Likhada; Dr. Rahulsinh B. Chauhan; Chaudhari Manisha Maheshchandra; Gauravkumar P Rathod; and Dehlvi Mohammadnaved J.
The application for the patent was published on May 1, under issue no. 18/2026.
According to the abstract released by the Intellectual Property India: "The present disclosure relates to a system and method for automatically extracting features and optimizing learning in machine learning environments. The system has a data acquisition module that is set up to gather different types of data from different sources, such as databases, sensors, and cloud platforms. A preprocessing module takes the data that has been collected and turns it into structured input representations. A data representation engine turns the processed data into computational feature embeddings. A self-evolving feature synthesis module automatically creates candidate features by finding relationships and dependencies between existing attributes. A multi-modal feature extraction processor gets useful features from different types of data, such as text, images, and time-series data. A feature optimization engine picks the best feature subsets to improve predictive performance and cut down on redundancy. The system also has a meta-learning optimization controller that can change learning parameters on the fly based on performance metrics. A model evaluation module looks at how accurate predictions are and how important features are to give feedback for ongoing improvement. This makes it possible to optimize machine learning models quickly and easily."
Disclaimer: Curated by HT Syndication.
