WEIGHTED NON-CROSSING QUANTILE REGRESSIONS

In this article we are concerned with a collection of multiple linear regressions that enable the researcher to gain an impression of the entire conditional distribution of a response variable given the specications for the explanatory variables. In particular, we investigate the advantage of using a new method of parametric estimation for non-crossing quantile regressions. The main tool is a weighting system of the observations that aims to reduce the eect of contamination of the sampled population on the estimated parameters by diminishing the eect of outliers. The performance of the new estimators has been evaluated on a number of data sets. We had considerable success with avoiding intersections and in the same time improving the global tting of conditional quantile regressions. We conjecture that in other situations (e.g. data with high level of skewness, non-constant variances, unusual and uncertain data) the method of weighted non-crossing quantiles will lead to estimators with good robustness properties.