MUMBAI, India, Aug. 17 -- Intellectual Property India has published a patent application (202641095542 A) filed by Gorige Sangeetha; and Dr. M. Janga Reddy on August 06, 2026, for System And Method For Dynamic Causal Fairness Correction In Streaming Machine Learning Systems.
Inventors include M. Sri Rama Lakshmi Reddy; Mohmad Ahmed Ali; Gunda Hariteja; Swetha Koduri; A. Prakash; and Puli Vijay Kumar.
The application for the patent was published on August 14, 2026, under issue no. 33/2026.
Abstract: A system (100) and method for dynamic causal fairness correction in streaming machine learning systems are disclosed. The system (100) comprises a memory (102), a processor (104), and a communication module (106). The communication module (106) receives streaming input data, inference telemetry, contextual environment information, and protected attribute indicators associated with one or more machine learning pipelines. A causal disparity detection module (108) dynamically constructs evolving causal dependency representations and identifies discriminatory causal propagation pathways across streaming inference events. A drift-adaptive fairness correction module (110) selectively applies adaptive fairness correction operations by modifying feature contribution weights, decision constraints, and causal influence relationships without requiring complete retraining of associated machine learning models. A streaming policy orchestration module (112) continuously generates context-adaptive fairness enforcement policies based on real-time drift behavior and synchronizes fairness correction operations across distributed streaming environments.
Disclaimer: Curated by HT Syndication.
