Open gdevos010 opened 2 years ago
Is your feature request related to a current problem? Please describe. Multivariate Quantile Function Forecaster
The MQF2 distribution loss looks to boost forecasting performance.
That could perhaps take the form of a new Likelihood object to be placed on top of our existing neural nets (we need to see what's the difference / delta w.r.t. our current QuantileRegression though)
QuantileRegression
Is your feature request related to a current problem? Please describe. Multivariate Quantile Function Forecaster
The MQF2 distribution loss looks to boost forecasting performance.