D-optimal designs for multinomial logistic models
From MaRDI portal
Publication:2196203
DOI10.1214/19-AOS1834zbMATH Open1453.62606arXiv1707.03063OpenAlexW3030709430MaRDI QIDQ2196203FDOQ2196203
Authors: Xianwei Bu, Dibyen Majumdar, Jie Yang 
Publication date: 28 August 2020
Published in: The Annals of Statistics (Search for Journal in Brave)
Abstract: We consider optimal designs for general multinomial logistic models, which cover baseline-category, cumulative, adjacent-categories, and continuation-ratio logit models, with proportional odds, non-proportional odds, or partial proportional odds assumption. We derive the corresponding Fisher information matrices in three different forms to facilitate their calculations, determine the conditions for their positive definiteness, and search for optimal designs. We conclude that, unlike the designs for binary responses, a feasible design for a multinomial logistic model may contain less experimental settings than parameters, which is of practical significance. We also conclude that even for a minimally supported design, a uniform allocation, which is typically used in practice, is not optimal in general for a multinomial logistic model. We develop efficient algorithms for searching D-optimal designs. Using examples based on real experiments, we show that the efficiency of an experiment can be significantly improved if our designs are adopted.
Full work available at URL: https://arxiv.org/abs/1707.03063
Recommendations
- Optimum Experimental Designs for Multinomial Logistic Models
- Optimal designs for generalized linear models with multiple design variables
- Optimal designs for binary response models with multiple nonnegative variables
- Optimal design for discrete choice experiments
- D-optimal designs with ordered categorical data
approximate designFisher information matrixexact designmultinomial responseminimally supported designlift-one algorithm
Statistical aspects of information-theoretic topics (62B10) Generalized linear models (logistic models) (62J12) Optimal statistical designs (62K05)
Cites Work
- Vector generalized linear and additive models. With an implementation in R
- An Introduction to Generalized Linear Models, Fourth Edition
- Title not available (Why is that?)
- Bayesian experimental design: A review
- Title not available (Why is that?)
- Title not available (Why is that?)
- On Optimal Designs for Nonlinear Models: A General and Efficient Algorithm
- General equivalence theory for optimum designs (approximate theory)
- Optimum experimental designs, with SAS
- Analysis of ordinal categorical data.
- Title not available (Why is that?)
- Title not available (Why is that?)
- Partial Proportional Odds Models for Ordinal Response Variables
- A review of ordinal regression models applied on health-related quality of life assessments
- Title not available (Why is that?)
- The analysis of ordered categorical data: An overview and a survey of recent developments. (With discussion)
- Optimal design for nonlinear response models
- Title not available (Why is that?)
- Design issues for generalized linear models: a review
- Optimum Experimental Designs for Multinomial Logistic Models
- Parameter estimation problems with singular information matrices
- Optimal design for the proportional odds model
- Optimal designs for 2k factorial experiments with binary response
- D-optimal Factorial Designs under Generalized Linear Models
- D-optimal Designs with Ordered Categorical Data
Cited In (15)
- Locally D-Optimal Designs for Hierarchical Response Experiments
- An algorithmic approach to construct D-optimal saturated designs for logistic model
- Locally D-Optimal Design for a Logit Model in Discrete Choice Experiment
- Optimal designs for beta-binomial logistic regression models
- Optimum Experimental Designs for Multinomial Logistic Models
- Optimal designs and limiting optimal designs for a trinomial response
- A note on the construction of locally \(D\)- and \(D_S\)-optimal designs for the binary logistic model with several explanatory variables
- ForLion: a new algorithm for D-optimal designs under general parametric statistical models with mixed factors
- \(D\)- and \(A\)-optimal allocations in a multipoint grouped logit designed experiment
- Highly Efficient Designs for Logistic Models with Categorical Variables
- \(D\)-optimal designs for the two-variable binary logistic regression model with interaction
- Optimal Experimental Designs for Multilevel Logistic Models with Two Binary Predictors
- Identifying the most appropriate order for categorical responses
- D-optimal designs for logistic models with three and four parameters
- \(D\)-optimal conjoint choice designs with no-choice options for a nested logit model
Uses Software
This page was built for publication: D-optimal designs for multinomial logistic models
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2196203)
