The students’ academic performance prediction is an important task since it helps the teachers to know about the predicted mark of students in the end semester. It helps them to identify the students who need additional help and counselling so it can lead to reduced students’ failure rate in the final examination. In this research, we have made an attempt to predict the student’s academic performance in the end semester examination by considering three Continuous Assessment marks of them in various subjects. The marks are predicted using various regression algorithms for the same dataset and their accuracies are compared to determine the best algorithm in meticulous prediction of end semester marks. We have used Multiple Linear Regression, K-Nearest Neighbour regression (KNNR), Random Forest Regression (RFR), Gradient Boosted Regression (GBR) and Extreme Gradient Boosting (XGB), AdaBoost Regression, Bagging-meta-estimator. In these analysis, Gradient Boosted Regression provides high accuracy and Bagging-meta-estimator provides low accuracy.