LeratoTheko/Tax_Prediction

A Machine Learning Model to Predict Revenue Services Lesotho Tax **Overview** This project aims to build a machine-learning model that can predict revenue values for different revenue categories in a given year for Revenue Services Lesotho. The model is developed using XGBoost, a powerful gradient-boosting library, to achieve accurate predictions. **Project Objective** The main objective of this project is to build a predictive model that can assist Revenue Services Lesotho in forecasting revenue values for various categories (e.g., tatal and non-total revenue, duties, fees) in a specific year. By accurately predicting revenue, the organization can better plan its budget, allocate resources effectively, and optimize revenue collection strategies. **Project Structure** Data: Contains dataset Data RSL.csv containing the revenue data, the dataset is used to train the model. Notebooks: Includes Revenue Services Lesotho Tax Prediction.ipynb, a Jupyter notebook containing data preprocessing and model training code. Models: Contains the trained XGBoost model RSL_Model.model saved as model file. **Usage** To use the trained model for revenue prediction, follow these steps: Ensure you have the required dataset in CSV format. If not, replace the existing RSL Data.csv file in the data directory with your dataset. Open and run the Revenue Services Lesotho Tax Prediction.ipynb notebook located in the notebooks directory. This notebook contains detailed instructions and code for data preprocessing, model training, and evaluation. After running the notebook, the trained XGBoost model will be saved as RSL_Model.model in the models directory. You can now use the trained model for revenue prediction by loading it in your Python script or Jupyter notebook and providing input data. **Contributers** github.com github.com github.com github.com