Methods to Deal with Unknown Populational Minima during Parameter Inference

TL;DR

This paper introduces six methods to address the challenge of unknown population minima in parameter inference, especially useful for multiple datasets, combining theoretical rigor and simplicity for practical application.

Contribution

It proposes six novel methods to handle unknown population minima efficiently, including two with full theoretical support and four simple, theoretically grounded approaches.

Findings

Two methods have full theoretical support but are complex.

Four simple methods are effective and based on non-parametric results.

Methods improve likelihood maximization and enable parameter grid reuse.

Abstract

There is a myriad of phenomena that are better modelled with semi-infinite distribution families, many of which are studied in survival analysis. When performing inference, lack of knowledge of the populational minimum becomes a problem, which can be dealt with by making a good guess thereof, or by handcrafting a grid of initial parameters that will be useful for that particular problem. These solutions are fine when analyzing a single set of samples, but it becomes unfeasible when there are multiple datasets and a case-by-case analysis would be too time consuming. In this paper we propose methods to deal with the populational minimum in algorithmic, efficient and/or simple ways. Six methods are presented and analyzed, two of which have full theoretical support, but lack simplicity. The other four are simple and have some theoretical grounds in non-parametric results such as the law of iterated logarithm, and they exhibited very good results when it comes to maximizing likelihood and being able to recycle the grid of initial parameters among the datasets. With our results, we hope to ease the inference process for practitioners, and expect that these methods will eventually be included in software packages themselves.