A survey of a hurdle model for heavy-tailed data based on the generalized lambda distribution

TL;DR

This paper surveys the use of the Generalized Lambda Distribution (GLD) in hurdle models for heavy-tailed data with excess zeros, demonstrating superior performance over other models on healthcare expense data.

Contribution

It introduces a flexible hurdle model based on the GLD for heavy-tailed, zero-inflated data, and compares its effectiveness with existing models.

Findings

GLD-based hurdle models outperform GPD models on healthcare data.

The proposed models effectively handle heavy tails and excess zeros.

Empirical results show better fit with GLD models.

Abstract

In this survey we present an extensive research of the vast literature about the Generalized Lambda Distribution (GLD) and propose a hurdle, or two-way, model whose associated distribution is the GLD in order to meet the demand for a highly flexible model of heavy-tailed data with excess of zeros. We apply the developed models to a dataset consisting of yearly healthcare expenses, a typical example of heavy-tailed data with excess of zeros. The fitted models are compared with models based on the Generalised Pareto Distribution and it is established that the GLD models perform best.