Mohsin64/building-energy-predictor

Building Energy Efficiency Predictor

Predict the heating load of buildings based on architectural characteristics using machine learning.

Features

• —9 Trained Models: Ridge, Lasso, and ElasticNet with polynomial degrees 1-3
• —Best Model: Lasso (Degree 3) with R² = 0.99
• —Interactive Interface: Adjust building parameters with sliders
• —Model Comparison: Compare predictions across all models

How to Use

1. 1.Select a Model from the dropdown
2. 2.Adjust Building Parameters using the sliders
3. 3.Click "Predict Heating Load"
4. 4.View the prediction and energy efficiency rating

Dataset

UCI Machine Learning Repository - Energy Efficiency Dataset (768 buildings)

Technical Details

• —Polynomial feature engineering
• —StandardScaler + OneHotEncoder preprocessing
• —5-Fold Cross-Validation
• —Regularized regression (L1, L2, ElasticNet)