Dynamic Mixture of Experts Models for Online Prediction
- 2 December 2022
- journal article
- research article
- Published by Taylor & Francis in Technometrics
- Vol. 65 (2) , 257-268
Abstract
A mixture of experts models the conditional density of a response variable using a mixture of regression models with covariate-dependent mixture weights. We extend the finite mixture of experts model by allowing the parameters in both the mixture components and the weights to evolve in time by following random walk processes. Inference for time-varying parameters in richly parameterized mixture of experts models is challenging. We propose a sequential Monte Carlo algorithm for online inference and based on a tailored proposal distribution built on ideas from linear Bayes methods and the EM algorithm. The method gives a unified treatment for mixtures with time-varying parameters, including the special case of static parameters. We assess the properties of the method on simulated data and on industrial data where the aim is to predict software faults in a continuously upgraded large-scale software project.Keywords
All Related Versions
This publication has 50 references indexed in Scilit:
- Efficient Block Sampling Strategies for Sequential Monte Carlo MethodsJournal of Computational and Graphical Statistics, 2006
- Modeling nonlinear time series with local mixtures of generalized linear modelsCanadian Journal of Statistics, 2005
- Mixtures-of-Experts of Autoregressive Time Series: Asymptotic Normality and Model SpecificationIEEE Transactions on Neural Networks, 2005
- Central limit theorem for sequential Monte Carlo methods and its application to Bayesian inferenceThe Annals of Statistics, 2004
- Dealing With Label Switching in Mixture ModelsJournal of the Royal Statistical Society Series B: Statistical Methodology, 2000
- Filtering via Simulation: Auxiliary Particle FiltersJournal of the American Statistical Association, 1999
- Sequential Monte Carlo Methods for Dynamic SystemsJournal of the American Statistical Association, 1998
- Novel approach to nonlinear/non-Gaussian Bayesian state estimationIEE Proceedings F Radar and Signal Processing, 1993
- Dynamic Bayesian Models for Survival DataJournal of the Royal Statistical Society Series C: Applied Statistics, 1991
- Dynamic Generalized Linear Models and Bayesian ForecastingJournal of the American Statistical Association, 1985