conftrace_

2020 PGM PGM 2020

On a possibility of gradual model-learning

Abstract

In this paper, the term of gradual learning describes the process, in which an $n$-dimensional model is constructed in $n$ steps; each step increases the dimensionality of the constructed model by one. The approach is explained using the apparatus of compositional models since its algebraic properties seem to serve the purpose best. The paper shows also the equivalence of compositional models and Bayesian networks, and thus the paper gives a hint that the approach applies to the graphical model as well.

Authors

Radim Jiroušek

Topics

Artificial Intelligence > Bayesian & Probabilistic > Probabilistic Modeling Artificial Intelligence > Learning Paradigms > Transfer Learning Machine Learning > Core Methods > Representation Learning Machine Learning > Bayesian & Probabilistic > Bayesian Learning Machine Learning > Learning Paradigms > Curriculum Learning Machine Learning > Bayesian & Probabilistic > Bayesian Networks Artificial Intelligence > Bayesian & Probabilistic > Bayesian Networks

Keywords

bayesian network graphical model model learning gradual learning compositional model dimensional model

More from PGM 2020

All PGM 2020 papers →

Scalable Bayesian Network Structure Learning via Maximum Acyclic Subgraph

Kernel-based Approach for Learning Causal Graphs from Mixed Data

Lifted Query Answering in Gaussian Bayesian Networks

Causal Feature Learning for Utility-Maximizing Agents

Lifted Weight Learning of Markov Logic Networks (Revisited One More Time)

Prediction of High Risk of Deviations in Home Care Deliveries