This repository contains a couple of Jupyter Notebooks that explore the task of forecasting electricity consumption using six different machine learning models: fbprophet, XGBoost, Random Forest Regression, Linear Regression, Multi-Layer Perceptron (MLP), and Ridge Regression.
Accurate forecasting of electricity consumption is crucial for efficient energy management and planning. This project aims to compare the performance of various machine learning models in predicting electricity consumption, providing insights into the most suitable approach for this task.
The dataset used in this project is from Kaggle, and it is provided under License CC0: Public Domain.
The following machine learning models are implemented and evaluated in the notebooks:
- FBProphet: A forecasting library developed by Facebook's Core Data Science team that provides a simple and effective way to forecast time series data with trend and seasonality.
- XGBoost: A gradient boosting decision tree algorithm known for its high performance and efficiency.
- Random Forest Regression: An ensemble learning method that combines multiple decision trees to improve the accuracy of predictions.
- Linear Regression: A simple and widely-used regression technique that models the relationship between the input features and the target variable.
- Multi-Layer Perceptron (MLP): A type of artificial neural network that can learn complex non-linear relationships in the data.
- Ridge Regression: A variant of linear regression that introduces a penalty term to address the problem of multicollinearity.
To use the notebook, you will need to have Jupyter Notebook or JupyterLab installed on your system, along with the necessary Python libraries (e.g., NumPy, Pandas, Scikit-learn, XGBoost). You can then download or clone the repository and open the Jupyter Notebook file to explore the code and the results. Alternatively, you can just run it on Google Colab and replicate the notebooks for yourself.
The notebooks present the performance of each model in terms of various evaluation metrics, such as Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and R-squared. The results are discussed and compared to provide insights into the strengths and weaknesses of each approach.
If you find any issues or have suggestions for improvements, feel free to open an issue or submit a pull request. Contributions are welcome!
This project is licensed under the MIT License.