Predicting building energy consumption with the power of machine learning.
This project uses machine learning to predict building energy usage based on environmental and building-related features. The goal is to understand key drivers of consumption and support data‑driven energy efficiency decisions.
- Exploratory data analysis of energy consumption patterns.
- Feature engineering and data preprocessing for model training.
- Machine learning models for energy usage prediction (e.g., regression).
- Model evaluation with appropriate metrics and visualizations.
- Clear, reproducible workflow in a single Jupyter Notebook.
Given historical measurements and building attributes, the task is to build a model that can estimate future energy consumption. This helps facility managers and stakeholders identify inefficiencies and plan optimization strategies.
- Python
- Jupyter Notebook
- NumPy, Pandas
- Scikit-learn
- Matplotlib, Seaborn
- Clone the repository: -git clone https://github.com/rivu-intel45/building-energy-forecasting.git cd building-energy-forecasting
- Create and activate a virtual environment (optional but recommended).
- Install dependencies: -pip install -r requirements.txt
- Launch Jupyter and open the notebook:
-jupyter notebook
Then open
prediction-of-energy-consumption-using-ml.ipynb.
- Data loading and initial inspection
- Handling missing values and outliers
- Feature engineering and scaling
- Training and tuning machine learning models
- Performance evaluation and result interpretation
The dataset used in this project comes from Kaggle (building energy consumption). Please refer to the Kaggle page for licensing and download the data directly from there, then place it in the appropriate folder before running the notebook.
- Experiment with advanced models (e.g., gradient boosting, XGBoost, LightGBM).
- Hyperparameter tuning for better accuracy.
- Deployment of the trained model as an API or simple web app.