Students and teachers alike, want to enhance the academic performance (GPA - Grade Point Average). As part of the Artificial Intelligence (AI, CE4715) course at AU-VMES, students explored the correlation between academic performance (GPA) and behavioral/social features of the students using machine learning models. Raw data of 72 undergraduate students were collected then features were encoded and analyzed. 80% of the data was used for training and 20% for testing by three different machine learning models - i) Random Forest, ii) Linear Regression and iii) Decision Tree. Out of the three, Random Forest model showed the best output. Our study reflected that certain behaviors: year of study (1st-4th year), credits taken in each semester, relationship with lecturers show positive correlation whereas gaming (computer) hours, partying, no/minimum class attendance and assignment submission showed negative correlation. These results can be used as a feedback mechanism to enhance students’ behavioral patterns, thereby improving academic performance and overall well-being.