P1 Housing Price Prediction
Model
Keras regression model with:
- Dense(16, relu)
- Dense(8, relu)
- Dense(1, linear)
Loss:
Task
Predict regional housing price (mhv) from tabular features.
Input Features
['mi', 'ha', 'ar', 'pop', 'la', 'lg']
Preprocessing
- clipping on: ['mi', 'ar', 'pop']
log1p on: ['mi', 'ar', 'pop']StandardScaler
Validation Metrics
- RMSE: 0.5868
- MAE: 0.4005
- R²: 0.7492
Test Metrics
- RMSE: 0.5945
- MAE: 0.4039
- R²: 0.7303
Files
model.kerasscaler.joblibfeature_names.jsonpreprocess.jsonmetrics.json
Notes
This repository contains the trained model and preprocessing metadata.
