The dataset contains National Basketball Association’s tabular data of 22 columns, all regarding a player's performance records such as e.g. the number of 3 Points made. Upon applying Logistics Regression, the accuracy of the model is found to be 0.67. Precision recall tradeoff lies somewhere near 0.45 and the AUC-ROC curve covers an area of 0.74. Applying bagging, we get an accuracy score of 0.72 whereas applying Random Forest, the test score increases sharply till 0.75 and then remain almost similar. Hyper-parameter tuning of max-sample, we can see the test score increases till 0.78 and the reduces by little,tuning max-features we can see there is a significant difference between test score and train score which mean there is overfitting in the model. Regularization is to be performed to reduce the overfitting. The dataset is taken from kaggle: https://www.kaggle.com/sveneschlbeck/nba-players-career-duration