Parsimonious model selection using information theory: a modified selection rule

Ecology ◽  
2021 ◽  
Author(s):  
Luke A. Yates ◽  
Shane A. Richards ◽  
Barry W. Brook
Keyword(s):  
Information Theory ◽  
Model Selection ◽  
Selection Rule ◽  
Parsimonious Model

Ecological Model Selection via Evolutionary Computation and Information Theory

2004 ◽  
Vol 5 (2) ◽  
pp. 229-241 ◽  
Author(s):  
James P. Hoffmann ◽  
Christopher D. Ellingwood ◽  
Osei M. Bonsu ◽  
Daniel E. Bentil
Keyword(s):  
Information Theory ◽  
Model Selection ◽  
Evolutionary Computation ◽  
Ecological Model

scholarly journals Adaptive Estimation for Epidemic Renewal and Phylogenetic Skyline Models

2020 ◽  
Vol 69 (6) ◽  
pp. 1163-1179 ◽  
Author(s):  
Kris V Parag ◽  
Christl A Donnelly
Keyword(s):  
Information Theory ◽  
Model Selection ◽  
Adaptive Estimation ◽  
Minimum Description Length ◽  
Target Population ◽  
Theory Model ◽  
Information Criteria ◽  
Classification Problems ◽  
Effective Population ◽  
Incident Cases

Abstract Estimating temporal changes in a target population from phylogenetic or count data is an important problem in ecology and epidemiology. Reliable estimates can provide key insights into the climatic and biological drivers influencing the diversity or structure of that population and evidence hypotheses concerning its future growth or decline. In infectious disease applications, the individuals infected across an epidemic form the target population. The renewal model estimates the effective reproduction number, R, of the epidemic from counts of observed incident cases. The skyline model infers the effective population size, N, underlying a phylogeny of sequences sampled from that epidemic. Practically, R measures ongoing epidemic growth while N informs on historical caseload. While both models solve distinct problems, the reliability of their estimates depends on p-dimensional piecewise-constant functions. If p is misspecified, the model might underfit significant changes or overfit noise and promote a spurious understanding of the epidemic, which might misguide intervention policies or misinform forecasts. Surprisingly, no transparent yet principled approach for optimizing p exists. Usually, p is heuristically set, or obscurely controlled via complex algorithms. We present a computable and interpretable p-selection method based on the minimum description length (MDL) formalism of information theory. Unlike many standard model selection techniques, MDL accounts for the additional statistical complexity induced by how parameters interact. As a result, our method optimizes p so that R and N estimates properly and meaningfully adapt to available data. It also outperforms comparable Akaike and Bayesian information criteria on several classification problems, given minimal knowledge of the parameter space, and exposes statistical similarities among renewal, skyline, and other models in biology. Rigorous and interpretable model selection is necessary if trustworthy and justifiable conclusions are to be drawn from piecewise models. [Coalescent processes; epidemiology; information theory; model selection; phylodynamics; renewal models; skyline plots]

Information-theory-based model selection for determining the main vector and period of transmission of Potato virus Y

2011 ◽  
Vol 159 (3) ◽  
pp. 414-427 ◽  
Author(s):  
S.M. Kirchner ◽  
T.F. Döring ◽  
L.H. Hiltunen ◽  
E. Virtanen ◽  
J.P.T. Valkonen
Keyword(s):  
Information Theory ◽  
Model Selection ◽  
Potato Virus ◽  
Potato Virus Y ◽  
Selection For

Parsimonious Model Selection for Tissue Segmentation and Classification Applications: A Study Using Simulated and Experimental DTI Data

2007 ◽  
Vol 26 (11) ◽  
pp. 1576-1584 ◽  
Author(s):  
R.Z. Freidlin ◽  
E. Ozarslan ◽  
M.E. Komlosh ◽  
Lin-Ching Chang ◽  
Cheng Guan Koay ◽  
...  
Keyword(s):  
Model Selection ◽  
Tissue Segmentation ◽  
Selection For ◽  
Parsimonious Model

Information Theory and Log-Likelihood Models: A Basis for Model Selection and Inference

1998 ◽  
pp. 32-74 ◽  
Author(s):  
Kenneth P. Burnham ◽  
David R. Anderson
Keyword(s):  
Information Theory ◽  
Model Selection ◽  
Log Likelihood

A model selection rule for sinusoids in white Gaussian noise

1996 ◽  
Vol 44 (7) ◽  
pp. 1744-1751 ◽  
Author(s):  
P.M. Djuric
Keyword(s):  
Model Selection ◽  
Gaussian Noise ◽  
Selection Rule ◽  
White Gaussian Noise

scholarly journals Model Selection Using Information Theory and the MDL Principle

2004 ◽  
Vol 33 (2) ◽  
pp. 230-260 ◽  
Author(s):  
Robert A. Stine
Keyword(s):  
Information Theory ◽  
Model Selection ◽  
Mdl Principle

Parsimonious model selection for tissue classification: a DTI study of zebrafish

2007 ◽  
Author(s):  
Raisa Z. Freidlin ◽  
Michal E. Komlosh ◽  
Murray H. Loew ◽  
Peter J. Basser
Keyword(s):  
Model Selection ◽  
Tissue Classification ◽  
Selection For ◽  
Parsimonious Model

scholarly journals Model Uncertainty and Exchange Rate Forecasting

2017 ◽  
Vol 52 (1) ◽  
pp. 341-363 ◽  
Author(s):  
Roy Kouwenberg ◽  
Agnieszka Markiewicz ◽  
Ralph Verhoeks ◽  
Remco C. J. Zwinkels
Keyword(s):  
Exchange Rate ◽  
Model Selection ◽  
Model Uncertainty ◽  
Selection Rule ◽  
Time Varying ◽  
Exchange Rate Forecasting ◽  
Out Of Sample ◽  
Small Set ◽  
Strong Performance ◽  
Exchange Rate Models

Exchange rate models with uncertain and incomplete information predict that investors focus on a small set of fundamentals that changes frequently over time. We design a model selection rule that captures the current set of fundamentals that best predicts the exchange rate. Out-of-sample tests show that the forecasts made by this rule significantly beat a random walk for 5 out of 10 currencies. Furthermore, the currency forecasts generate meaningful investment profits. We demonstrate that the strong performance of the model selection rule is driven by time-varying weights attached to a small set of fundamentals, in line with theory.

Model Selection

2021 ◽  
pp. 149-164
Author(s):  
Timothy E. Essington
Keyword(s):  
Information Theory ◽  
Model Selection ◽  
Goodness Of Fit ◽  
Danaus Plexippus ◽  
Predictive Ability ◽  
Information Criterion ◽  
Monarch Butterfly ◽  
Evaluation Framework ◽  
Alternative Hypotheses ◽  
Types Of Information

The chapter “Model Selection” provides a brief overview of alternative ways of thinking about what is (are) the best model(s) and shows the core motivation behind the Akaike information criterion as a measure of the predictive ability of a model fitted via maximum likelihood. It then gives stepwise practical guidance for using information theory as a basis of model selection, including nested versus non-nested models, goodness of fit, and overdispersion. Advanced topics cover some of the philosophy of information theory, other types of information theory criteria, and other ways of evaluating the predictive ability of models. As an example, the chapter examines the case of the Western monarch butterfly (Danaus plexippus plexippus), which, over the past two decades, has experienced a 97% decline from its historical average abundance, declining by 86% from 2017 to 2018 alone. Undoubtedly, there is more than one cause—indeed, overwintering habitat loss and pesticide use are both believed to be important contributors to the decline. Adopting a hypothesis-evaluation framework makes it possible to consider multiple alternative hypotheses simultaneously and measure degrees of support for alternative hypotheses.

Sign in / Sign up

Export Citation Format

Share Document