This project demonstrates time series forecasting using machine learning models like XGBoost and LightGBM. Built using Python and visualized with Matplotlib and Seaborn, it provides insights into trends and predictions using historical time series data.
- Forecasting using XGBoost and LightGBM
- Visualization of original and predicted time series
- Performance evaluation using MSE and other metrics
- Easily extensible for other datasets or models
-
Python
-
Pandas for data manipulation and preprocessing
-
NumPy for numerical computations
-
Matplotlib & Seaborn for data visualization
-
Scikit-learn for train-test splitting and evaluation
-
XGBoost and LightGBM for forecasting models
Time-Series-Forecasting-Dashboard
├── Time_Series_Forecasting.ipynb # Jupyter Notebook for training and evaluating models
├── requirements.txt # Required Python libraries
└── README.md # Project overview and instructions
- Clone the repository and navigate to it on your machine
git clone https://github.com/Im-Arth1307/Time-Series-Forecasting-Dashboard.git
cd Time-Series-Forecasting-Dashboard- Install all the necessary dependencies
pip install -r requirements.txt
- Open Time_Series_Forecasting.ipynb in Jupyter Notebook or Jupyter Lab and run the cells sequentially.
Visualization of Feature Importance:
Dashboard:
Predicted Sales versus Actual Sales Graph:
Mean Square Precentage Error for Evaluation:
ML models used:
-
Add support for more time series datasets
-
Hyperparameter tuning and feature engineering
Contributions are always welcome ! Feel free to fork the repo, make improvements and create a pull request.
For any questions or suggestions, reach out to:
Atharva Jakhetiya
Gmail: jakhetiyaathava@gmail.com
LinkedIn: https://www.linkedin.com/in/atharva-jakhetiya/
GitHub: https://github.com/Im-Arth1307