MUMBAI, India, Aug. 12 -- Intellectual Property India has published a patent application (202621075934 A) filed by Symbiosis International Deemed University on June 18, 2026, for Explainable Scholarship Eligibility Prediction System Using Machine Learning And Shap-Based Interpretability.
Inventors include Prof. Rohit Pawar; Shruti Bawankar; Ayushi Ray; and Sanvi Wadhankar.
The application for the patent was published on August 07, 2026, under issue no. 32/2026.
Abstract: ABSTRACT EXPLAINABLE SCHOLARSHIP ELIGIBILITY PREDICTION SYSTEM USING MACHINE LEARNING AND SHAP-BASED INTERPRETABILITY The present invention discloses an explainable scholarship eligibility prediction system (100) that integrates machine learning with interpretable artificial intelligence for automated and transparent student financial aid determination. The system (100) comprises a frontend interface module (110) for data input and result visualization, a backend server module (130) implemented as a Flask REST API, a rule-based eligibility scoring module (140) that computes weighted composite scores from academic, socio-economic, and demographic factors, a Random Forest classifier module (150) trained on rule-generated labels achieving approximately 98 percent accuracy, and a SHAP-based explainability engine (160) that generates per-feature Shapley values and human-readable narrative explanations for each prediction. The hybrid framework bridges rigid policy rules with adaptive machine learning generalization while ensuring every automated decision is transparent and auditable. The system supports single-student real-time predictions and batch CSV processing through a RESTful API architecture. [
Disclaimer: Curated by HT Syndication.
