MUMBAI, India, Aug. 24 -- Intellectual Property India has published a patent application (202621082034 A) filed by Symbiosis International Deemed University on July 02, 2026, for Mlops-Integrated Ensemble Learning System For Real-Time Air Quality Index Prediction.
Inventors include Sahil Goyal; Vaibhav Sharma; Vivek Hotchandani; Girish Mahale; Mayur Gaikwad; and Dr. Aniket K. Shahade.
The application for the patent was published on August 21, 2026, under issue no. 34/2026.
Abstract: ABSTRACT MLOPS-INTEGRATED ENSEMBLE LEARNING SYSTEM FOR REAL-TIME AIR QUALITY INDEX PREDICTION The present invention provides an MLOps-integrated ensemble learning system (100) for real-time air quality index prediction. The system (100) comprises a data ingestion module (110) with provider-specific API connectors (112), a preprocessing and feature engineering module (120) for temporal standardization, missing value imputation, anomaly detection, and feature generation, a model training module (130) combining Random Forest (132) and XGBoost (134) regressors through a weighted ensemble combiner (136), an experiment tracking module (140) with MLflow-based tracking and model registry, an inference service module (150) with FastAPI-based prediction and health check endpoints, a containerization module (160) using multi-stage Docker containers, a CI/CD and automated retraining module (170), and a data version control module (180). The system provides end-to-end operational lifecycle management with automated retraining, drift detection (194), and rollback capabilities (196) for production-grade environmental monitoring. [
Disclaimer: Curated by HT Syndication.
