Data Center Energy Optimization System

An AI-powered data center energy optimization system that uses machine learning to predict energy consumption and provide real-time recommendations.

Features

• Machine Learning Models: Trained Random Forest model for energy prediction (MAE: 51.20 kWh)
• Real-Time Predictions: Interactive dashboard with live energy consumption forecasts
• Smart Recommendations: Intelligent alerts based on energy levels, cooling efficiency, and server load
• Interactive Visualizations: Historical data analysis with Plotly charts
• User-Friendly Interface: Streamlit-based dashboard with intuitive controls

Tech Stack

• Python 3.11
• Machine Learning: scikit-learn, XGBoost
• Data Processing: Pandas, NumPy, SciPy
• Visualization: Plotly, Streamlit
• Model: Random Forest Regressor

View

Go to https://adam-caio.streamlit.app/

Quick Start

See RUN.md for detailed instructions on running the project locally.

Project Structure

├── step1_dataset.py          # Dataset generation
├── step2_clean.py            # Data cleaning preprocessin
├── step3_models.py           # ML model training
├── step4_recommend.py        # Recommendation system
├── dashboard.py              # Streamlit dashboard
├── datacenter_data.csv       # Raw dataset
├── cleaned_data.csv          # Processed dataset
├── best_model.pkl            # Trained ML model
├── scaler.pkl                # Feature scaler
└── requirements.txt          # Python dependencies

Model Performance

• Algorithm: Random Forest Regressor
• Mean Absolute Error (MAE): 51.20 kWh
• Root Mean Squared Error (RMSE): 61.00 kWh
• Features: 6 (Server Load, Ambient Temperature, Cooling Efficiency, Hour, Day of Week, Month)
• Training Samples: 500

Development Process

See STEPS.md for a detailed breakdown of the development steps.

License

MIT License

Author

Adam Ahmed