Open saattrupdan opened 3 years ago
Dan, you might be interested in this link
Not sure if It is the right place for this message.
Since you mentioned Future work, you want to add support for neural networks. I would like to recommend looking into this paper: High-Quality Prediction Interval.
I started this notebook (WIP): https://github.com/leandroohf/machine_learning_algorithms/blob/master/dev/intro_to_prediction_interval.ipynb
and while doing my research, I discovered your packet doubt and decided to try.
Conformal Quantile Regression was introduced in Romano, Patterson & Candès and is a variant of quantile regression which calibrates the prediction intervals, yielding narrower intervals, while preserving theoretical coverage guarantees.
This could potentially be built into
QuantileLinearRegression
via aconformal
argument.