This Project is a Machine Learning Project which aims to Predict the score of the team based on certain parameters. The project uses a dataset containing ball-to-ball information from Season-1 to Season-10. download the dataset here. This Project adapts various Regression Algorithms to predict the score.
The data is preprocessed by removing unnecessary columns (manually), followed by removing the teams which do not play frequently and then converting them into numerical variables with the help of string indexing followed by one-hot encoding. and then followed by applying principal component analysis to check if there are any unwanted columns (low variance).
This involved implementing four regression models (Linear, Decision tree, Random forest and Gbt Regressor). Cross-validator is used along with hypertuning of the parameters to avoid overfitting. and then they were stacked together once with GBT and with Linear Regression.
All the models were evaluated with the help of RMSE(Root Mean Square Error). Out of them GBT Regressor Performed very well before stacking and after stacking as well.