MUMBAI, India, Oct. 26 -- Intellectual Property India has published a patent application (202611111765 A) filed by Ims Engineering College, Ghaziabad on September 17, 2026, for A Hybrid Machine Learning Based Transaction Security Method For Detecting Financial Fraud In Electronic Financial Networks.
Inventors include Prof. Dr. Anil K. Ahlawat; Dr. Vikash Goel; and Dr. Nishant Anand.
The application for the patent was published on October 23, 2026, under issue no. 43/2026.
Abstract: [044] The present invention discloses a hybrid machine learning based transaction security method for detecting financial fraud in electronic financial networks. The method comprises receiving a transaction data packet through a transaction security gateway, generating a privacy-preserving transaction representation, and producing a feature vector including transaction, behavioural, device, network, temporal, merchant, beneficiary, and relationship attributes. The feature vector is processed through a hybrid inference architecture including deterministic validation, supervised fraud classification, unsupervised anomaly detection, graph relationship risk analysis, and sequential behaviour risk analysis. Partial risk outputs and confidence indicators generated by the inference layers are dynamically fused by an adaptive fusion engine to generate a calibrated fraud risk score and fraud reason codes. A machine-readable transaction security control signal is transmitted to a transaction processing system to approve, decline, hold, review, or authenticate the transaction. Verified outcomes are used for feedback-based recalibration while preserving sensitive identifiers. Accompanied Drawing [FIGS. 1-2]
Disclaimer: Curated by HT Syndication.
