Bi-Level Path Following for Cross Validated Solution of Kernel Quantile Regression

Modeling of conditional quantiles requires specification of the quantile being estimated and can thus be viewed as a parameterized predictive modeling problem. Quantile loss is typically used, and it