Model Selection Techniques: An Overview

By Jie Ding, Vahid Tarokh, Yuhong Yang. 2018

343 citations recorded by the source corpus.

In the era of big data, analysts usually explore various statistical models or machine-learning methods for observed data to facilitate scientific discoveries or gain predictive power. Whatever data and fitting procedures are employed, a crucial step is to select the most appropriate model or method from a set of candidates. Model selection is a key ingredient in data analysis for reliable and reproducible statistical inference or prediction, and thus it is central to scientific studies in such fields as ecology, economics, engineering, finance, political science, biology, and epidemiology. There has been a long history of model selection techniques that arise from researches in statistics, information theory, and signal processing. A considerable number of methods has been proposed, following different philosophies and exhibiting varying performances. The purpose of this article is to provide a comprehensive overview of them, in terms of their motivation, large sample performance, and applicability. We provide integrated and practically relevant discussions on theoretical properties of state-of-the-art model selection approaches. We also share our thoughts on some controversial views on the practice of model selection.

Subfields: Artificial Intelligence, Statistics and Probability

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Flávio du Pin Calmon
2. Charles Marx
3. Berk Ustun
4. Cynthia Rudin
5. Lalitha Sankar
6. R. Caruana

In Statistics and Probability

1. Flávio du Pin Calmon
2. Charles Marx
3. Berk Ustun
4. Cynthia Rudin
5. Nadia Burkart
6. Marco F. Huber
7. Tong Wang

Citations in the corpus

• Predictive Multiplicity in Classification (2019, background): “Classical approaches for model selection choose a single model that is “closest” to a “true model” by breaking ties on the basis of measures such as the AIC, BIC, or K-CV error (see e.g., McAllister, 2007; Ding et al., 2018).”