BBAU Academic Admission Intelligence: An AI-Powered Decision Support System for Course Sustainability and Admission Optimization

Abstract

Public universities in India face unprecedented pressure to balance multiple competing constraints: rising demographic demand for higher education, constitutional equity mandates, fixed infrastructure capacity, and the policy demand for evidence-based governance. This paper presents BBAU Academic Admission Intelligence (BBAUI), an integrated AI-powered decision support system designed to address these challenges in real-time. The system combines machine learning for forecasting (Random Forest, XGBoost, ARIMA, Prophet), classical operations research for optimization (Linear Programming), and explainability frameworks (SHAP) to deliver strategic insights at the course level. Trained and validated on ten years of historical data from 106 courses across 23 departments, the system achieves forecasting accuracy exceeding 0.78 R-squared on held-out test data. The implementation spans 36 interactive analytical pages organized into four capability layers: descriptive analytics (dashboards and KPIs), predictive analytics (enrolment and outcome forecasting), prescriptive analytics (optimization recommendations), and equity auditing (reservation compliance and diversity monitoring). Results demonstrate that the data BBAU already collects is sufficient to predict next-year enrolment with acceptable accuracy, identify at-risk courses with explainable root causes, and recommend evidence-based interventions. The system runs on standard office hardware without requiring a database server, making it immediately deployable in resource-constrained academic settings. This research contributes to the emerging field of higher-education analytics in the Indian central-university context, where comparable decision-support systems remain rare.

Introduction

Higher education in India faces a dual challenge of rapidly increasing student demand and strict reservation-based equity requirements, while infrastructure such as faculty, labs, and seats remains fixed. This leads to complex and often subjective decision-making in admissions, course planning, and resource allocation. To address this, the proposed work introduces BBAU Academic Admission Intelligence (BBAUI), an AI-powered decision support system that integrates forecasting, optimization, equity auditing, and explainable analytics into a single platform.

The system is motivated by issues such as seat vacancies, inefficient resource use, difficulty in manual equity auditing, and the need for evidence-based governance as required by NEP-2020. It aims to answer key institutional questions about current status, future trends, causes of problems, and recommended actions using data-driven methods instead of intuition.

BBAUI uses multiple machine learning and statistical techniques including Random Forest, XGBoost, ARIMA, Prophet, clustering, survival analysis, and linear programming. It evaluates programme sustainability, forecasts enrolment, audits equity across nine demographic dimensions, and optimizes seat allocation under institutional constraints. Explainability is ensured using SHAP to make model decisions transparent and auditable.

The system is built as a Streamlit-based web application with 36 analytical modules, covering forecasting, equity, optimization, and data management. It processes ten years of academic data across 106 courses and includes features like scenario simulation, geographic analysis, and PDF reporting.

Results show strong forecasting performance, with ensemble models outperforming baselines and achieving high accuracy. Overall, BBAUI provides a scalable, transparent, and policy-aligned framework to improve admission planning, reduce vacancies, and support fair and data-driven decision-making in higher education.

Conclusion

This paper presents BBAU Academic Admission Intelligence, an integrated AI-powered decision support system for course-level analytics in a central Indian university. The system combines forecasting, optimization, equity auditing and explainability into a unified platform that runs on standard office hardware. Quantitative results show that machine-learning models beat target accuracy thresholds, while qualitative feedback from departmental coordinators confirms that the system produces actionable insights. The principal insight is that the data BBAU already collects is sufficient—without acquiring external feeds or new sensors—to forecast enrolment with acceptable accuracy, identify at-risk courses with interpretable root causes, and recommend evidence-based interventions. The system is immediately deployable, maintainable and aligned with NEP-2020 governance mandates.

References

This paper presents BBAU Academic Admission Intelligence, an integrated AI-powered decision support system for course-level analytics in a central Indian university. The system combines forecasting, optimization, equity auditing and explainability into a unified platform that runs on standard office hardware. Quantitative results show that machine-learning models beat target accuracy thresholds, while qualitative feedback from departmental coordinators confirms that the system produces actionable insights. The principal insight is that the data BBAU already collects is sufficient—without acquiring external feeds or new sensors—to forecast enrolment with acceptable accuracy, identify at-risk courses with interpretable root causes, and recommend evidence-based interventions. The system is immediately deployable, maintainable and aligned with NEP-2020 governance mandates.

Copyright

Copyright © 2026 Abhishek Rauniyar, G. Nitheesh Kumar, Ashish Verma, Abhishek Kumar, Astha Singh, Aman Kumar Sharma, Ashutosh Kumar. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.