Title Supervised Bayesian classification methods of Gaussian Spatio-temporal data based on generative machine learning models /
Translation of Title Erdvės-laiko Gausinių duomenų prižiurimo Bajesinio klasifikavimo metodai, pagrįsti generatyviniais mašininio mokymosi modeliais.
Authors Karaliutė, Marta
DOI 10.15388/vu.thesis.520
Full Text Download
Pages 144
Keywords [eng] Spatio-temporal data ; Gaussian Random Field models ; Bayes discriminant function
Abstract [eng] When using supervised generative models in classification problems, the assumption of independence of observations is applied that the data is assumed to be modelled by a Gaussian Random Field. In order to solve the tasks of supervised Bayesian classification of observations that are described by Gaussian spatio-temporal data models using spatio-temporal contextual information, new classification methods are needed that would expand the application possibilities of the generative model. The research aims to develop novel supervised Bayesian classification methods of Gaussian spatio-temporal data based on generative machine learning models. Spatial contextual information is incorporated into the probability distribution of feature values and class labels. Proposed parameter estimation strategies and rules. Derived analytical expressions for the numerous parameter estimators reduces the intractability for the considered data model types. The proposed classifiers are investigated using simulated and real data, and compared using different class label functions. This work extends the application of supervised generative classification methods for features with distributions of spatial contextual information described by Gaussian Random Field models. A generative algorithm has been developed that allows solving classification problems for Gaussian data with spatial and temporal dependence.
Dissertation Institution Vilniaus universitetas.
Type Doctoral thesis
Language English
Publication date 2023