This repository contains code for a sports prediction model developed using Jupyter Notebook in Google Colab. The model predicts player ratings based on various attributes.
To get started with this project, follow these steps:
- Clone this repository to your local machine.
- Open the Jupyter Notebook file
Group12_SportsPrediction.ipynbin Google Colab or any Jupyter Notebook environment.
The notebook begins with data preparation and preprocessing steps:
- Mounting Google Drive: Mounting Google Drive to access the dataset stored in Google Drive.
- Loading the Dataset: Loading the dataset
players_21.csvfrom Google Drive into a Pandas DataFrame. - Data Cleaning: Removing columns with a high percentage of missing values and irrelevant columns like URLs, player pictures, etc.
- Imputing Missing Values: Imputing missing values in numerical columns using the mean and categorical columns using the mode.
- Feature Scaling: Scaling the numerical columns using StandardScaler.
- Feature Encoding: Encoding categorical columns using LabelEncoder.
Feature engineering involves selecting highly correlated features with the target variable (overall player rating) to improve model performance.
The notebook trains multiple regression models including RandomForestRegressor, XGBRegressor, and GradientBoostingRegressor. The models are evaluated using mean absolute error (MAE) and mean squared error (MSE).
The trained model is tested with a new dataset players_22.csv. The new dataset undergoes the same preprocessing steps as the training dataset, and the optimized XGBRegressor model is used for prediction.
The trained models are saved using pickle for future deployment. Additionally, the StandardScaler object used for scaling the features is saved using joblib.
- Python 3
- Jupyter Notebook
- Required Python libraries (Pandas, NumPy, Scikit-learn, XGBoost, joblib)
- Demonstration Video: https://drive.google.com/file/d/1eJfRgwqafL2AxYITxnGdiiIoSPOmvGHU/view?usp=sharing
- Application Link: https://player-ratingapp.streamlit.app