A hidden process regression model for functional data description. Application to curve discrimination

Abstract

A new approach for functional data description is proposed in this paper. It consists of a regression model with a discrete hidden logistic process which is adapted for modeling curves with abrupt or smooth regime changes. The model parameters are estimated in a maximum likelihood framework through a dedicated Expectation Maximization (EM) algorithm. From the proposed generative model, a curve discrimination rule is derived using the Maximum A Posteriori rule. The proposed model is evaluated using simulated curves and real world curves acquired during railway switch operations, by performing comparisons with the piecewise regression approach in terms of curve modeling and classification.

References in corpus (4)

• Support vector machine for functional data classification
• Functional Multi-Layer Perceptron: a Nonlinear Tool for Functional Data Analysis
• Time series modeling by a regression approach based on a latent process
• A regression model with a hidden logistic process for signal parametrization

Cited by in corpus (11)

• Model-based clustering and segmentation of time series with changes in regime
• Model-based functional mixture discriminant analysis with hidden process regression for curve classification
• Nonparametric Hierarchical Clustering of Functional Data
• Gaussian Mixture Regression model with logistic weights, a penalized maximum likelihood approach
• An Introduction to the Practical and Theoretical Aspects of Mixture-of-Experts Modeling
• Mixture model-based functional discriminant analysis for curve classification
• Functional Mixture Discriminant Analysis with hidden process regression for curve classification
• Bayesian mixtures of spatial spline regressions
• Conditional Latent Block Model: a Multivariate Time Series Clustering Approach for Autonomous Driving Validation
• Unsupervised learning of regression mixture models with unknown number of components
• Piecewise regression mixture for simultaneous functional data clustering and optimal segmentation