Abstract:Identifying pure components in mixtures is a common yet challenging problem. The associated unmixing process requires the pure components, also known as endmembers, to be sufficiently spectrally distinct. Even with this requirement met, extracting the endmembers from a single mixture is impossible; an ensemble of mixtures with sufficient diversity is needed. Several spectral unmixing approaches have been proposed, many of which are connected to hyperspectral imaging. However, most of them assume highly diverse collections of mixtures and extremely low-loss spectroscopic measurements. Additionally, non-Bayesian frameworks do not incorporate the uncertainty inherent in unmixing. We propose a probabilistic inference approach that explicitly incorporates noise and uncertainty, enabling us to unmix endmembers in collections of mixtures with limited diversity. We use a Bayesian mixture model to jointly extract endmember spectra and mixing parameters while explicitly modeling observation noise and the resulting inference uncertainties. We obtain approximate distributions over endmember coordinates for each set of observed spectra while remaining robust to inference biases from the lack of pure observations and presence of non-isotropic Gaussian noise. Access to reliable uncertainties on the unmixing solutions would enable robust solutions as well as informed decision making.

A Bagging and Boosting Based Convexly Combined Optimum Mixture Probabilistic Model

Model-based clustering and classification using mixtures of multivariate skewed power exponential distributions

Enhancing Particle Swarm Optimization via probabilistic models

Bayesian estimation and prediction for certain mixtures

Portfolio selection under possibilistic mean-variance utility and a SMO algorithm

Robust modeling of mixture probabilistic principal component analysis and process monitoring application

Data-Driven Robust Chance Constrained Problems: A Mixture Model Approach

On the Behavior of the Expectation-Maximization Algorithm for Mixture Models

The Infinite Student'S T-Mixture For Robust Modeling

Finding Optimally Robust Data Mixtures via Concave Maximization

Bayesian mixtures of common factor analyzers: Model, variational inference, and applications

Combining information from independent sources through confidence distributions

Insurance Loss Modeling with Gradient Tree-Boosted Mixture Models

Optimal Combination of SVM and Bayesian Density Model Using Dempster-Shafer Theory

An Efficient and Provable Approach for Mixture Proportion Estimation Using Linear Independence Assumption.

Probabilistic Mixture Model-Based Spectral Unmixing

Robust Fitting of Mixture Models using Weighted Complete Estimating Equations

Learning Mixtures of Ranking Models

A Generalized Finite Mixture Model for Asymmetric Probability Distribution Observations

Factor Model of Mixtures

An Effective Model Selection Criterion for Mixtures of Gaussian Processes.