MUMBAI, India, Aug. 12 -- Intellectual Property India has published a patent application (202641094157 A) filed by Presidency University on August 04, 2026, for A System And Method For Self-Adaptive Hybrid Ai-Based Data Fusion And Predictive Analytics.
Inventors include Kumari K; and Dr Badri H S.
The application for the patent was published on August 07, 2026, under issue no. 32/2026.
Abstract: ABSTRACT A System and Method for Self-Adaptive Hybrid AI-Based Data Fusion and Predictive Analytics The present invention relates to an autonomous data fusion and predictive big data analytics system and method using a self-adaptive hybrid artificial intelligence framework. The system comprises a data acquisition layer configured to collect heterogeneous structured, semi-structured, and unstructured datasets from multiple sources. A preprocessing layer performs cleaning, normalization, missing value handling, transformation, and dimensionality reduction. A feature extraction layer employs convolutional neural networks, recurrent neural networks, long short-term memory networks, and transformer models to derive modality-specific features. A multi-modal data fusion layer combines extracted features into a unified fused representation. A hybrid learning layer integrates deep neural networks with ensemble learning models including random forest and gradient boosting for prediction generation. A self-adaptive optimization layer dynamically updates model parameters using prediction loss. A predictive analytics layer outputs classification, anomaly detection, forecasting, recommendations, and decision-support insights.
Disclaimer: Curated by HT Syndication.
