laxmena
Improved the accuracy of all four machine learning models in the Graduate Admission project
SKlearn's train test split splits the dataset by assigning 25% of the dataset to testing data. It is a good practice to assign atleast 30% of the dataset to the testing data to avoid over-fitting.
Original Accuracy: | | Regression Model | Accuracy Score | Mean Absolute Error | Root Mean Squared Error | | 0 | Linear Regression | 73.31 | 0.0585182 | 0.0732795 | | 1 | Decision Tree Regression | 54.34 | 0.0698 | 0.0958436 | | 2 | Random Forest Regression | 69.96 | 0.06003 | 0.0777449 | | 3 | SVR Model | 56.06 | 0.0831241 | 0.0940217 |
New Accuracy: | | Regression Model | Accuracy Score | Mean Absolute Error | Root Mean Squared Error | | 0 | Linear Regression | 78.97 | 0.0479976 | 0.0664456 | | 1 | Decision Tree Regression | 55.35 | 0.07 | 0.0968332 | | 2 | Random Forest Regression | 76.54 | 0.0512125 | 0.0701824 | | 3 | SVR Model | 69.07 | 0.0656934 | 0.0805838 |
