Dipartimento di Scienze Statistiche

Quantile regression is an extension of the classical least squares methods for the estimation of conditional expectations, to the estimation of conditional quantile functions. In the seminar we will study the main characteristics of this technique, and we will also show an application to insurance ratemaking. Two-part models based on generalized linear models are widely used in insurance rate-making for predicting the expected loss. We will explore an alternative method based on quantile regression which provides more information about the loss distribution and can also be used for insurance underwriting.