Machine Learning Optimized Model Using Gradient Discent For Enginnernig Admission Cut-Off

The engineering admission process in India is governed by centralized entrance examinations such as the (MHT-CET) Maharashtra Common Entrance Test or (JEE) Joint Entrance Examination. The existing system depends on rule-based allocation mechanisms and static cutoff that lack predictive personalization, intelligence and transparency. This research proposes the Gradient Descent algorithm of Machine Learning for engineering admission prediction model. Logistic Regression is applied to find the probability of admission based on multiple attributes, like entrance examination score, reservation category, higher secondary (PCM) percentage, institutional preference ranking, demographic factors and historical cutoff trends. Experimental evaluation is performed using simulated candidate profiles and historical admission datasets.