In the last few years, the evaluation of students’ movement within the university campus has become a common criterion for whether students attend classes regularly or spend time in other Points of Interest Locations (POILs) in the field of education. Each student has a different interest based on their mentality. Once the students enter the university campus they visit different POILs like the cafeteria, Maggi Point, pizza shop, etc. The research considered based on the REVA University (RU) dataset as input to anticipate students’ Length of Stay (LOS) based on a student's Point of Interest. Based on this article proposes a Hybrid framework to predict the LOS of REVA University students based on the POIL of considerable importance. This research work has been carried out by considering machine learning (ML) algorithms named Linear Regression (LiR), Lasso regression (LR), Elastic Regression(ER), Ridge regression (RR), K-nearest Neighbour (KNN), Extreme Gradient Boosting Regression (XGBR), Random Forest Regression (RFR), Hybrid Regression Model (HRM) and three Categorical Encoding Methods named label encoding, one-hot encoding, target encoding, methods help in constructing the regression prediction model. The dataset used in the research comprises 18157 entries with various features, sourced from the Dataset of Student Location Prediction in 2023. The performance analysis of the proposed framework has been carried out by considering Mean squared error (MSE) and coefficient of determination (R2). The results show that the universal computing framework, which uses the Hybrid algorithm in conjunction with the suggested encoding method, achieves the greatest MSE score (0.18774) and R2 (0.8787). The purpose of this study is to determine if a student enters the university, attends classes regularly, or remains in other Point of Interest Locations (POILs) on campus will be analyzed and recorded.