A Review on Quantile Regression for Stochastic Computer Experiments
This work provides practical guidance for researchers and practitioners in computational fields using stochastic computer experiments, though it is incremental as it reviews and compares existing methods rather than introducing new ones.
The authors conducted an empirical study comparing six quantile regression metamodels across three categories for stochastic computer experiments, testing them on problems varying in training set size, input dimension, signal-to-noise ratio, and probability density at target quantiles. They found good contrasts among the metamodels, enabling pattern extraction and providing guidelines for users to select the best method for specific problems.
We report on an empirical study of the main strategies for quantile regression in the context of stochastic computer experiments. To ensure adequate diversity, six metamodels are presented, divided into three categories based on order statistics, functional approaches, and those of Bayesian inspiration. The metamodels are tested on several problems characterized by the size of the training set, the input dimension, the signal-to-noise ratio and the value of the probability density function at the targeted quantile. The metamodels studied reveal good contrasts in our set of experiments, enabling several patterns to be extracted. Based on our results, guidelines are proposed to allow users to select the best method for a given problem.