Regression¶
Predict house values in California districts from census data. Dataset: California housing.
Training¶
Make the model itself uncertainty-aware.
-
House prices with a mixture density network
Predict the full distribution of house prices, including the spike at the census price cap.
GaussianHead·MixtureGaussianHead
Post-processing¶
Add calibrated uncertainty to an already trained model, without retraining it.
-
Adaptive conformal intervals for house prices
Make conformal intervals as wide as each prediction is uncertain, and keep their 90% guarantee.
GaussianHead·SplitConformalRegressor