↓Skip to main content
Eliezer de Souza da Silva

Eliezer de Souza da Silva

Researcher in Probabilistic AI/ML, Statistics and Data Science

Recent

On Divergence Measures for Training GFlowNets

·276 words
Tiago Da Silva Eliezer De Souza Da Silva Diego Mesquita Publication Conference NeurIPS Deep Generative Models Statistical Divergences Variational Inference Generative Flow Networks

Novel approach to training Generative Flow Networks (GFlowNets) by minimizing divergence measures such as Renyi-$\alpha$, Tsallis-$\alpha$, and Kullback-Leibler (KL) divergences. Stochastic gradient estimators using variance reduction techniques leads to faster and stabler training.

Analyzing GFlowNets: Stability, Expressiveness, and Assessment

·288 words
Tiago Da Silva Eliezer De Souza Da Silva Rodrigo Barreto Alves Luiz Max Carvalho Amauri H Souza Samuel Kaski Vikas Garg Diego Mesquita Publication Preprint Workshop SPIGM@ICML Deep Generative Models Theoretical ML Generative Flow Networks

How balance violations impact the learned distribution, motivating an weighted balance loss to improve training. For graph distributions, there are scenarios where balance is unattainable, and richer embeddings of children’s states is needed enhance expressiveness. To measure of distributional correctness in GFN we introduce a provable correct novel assessment metric.

Human-in-the-Loop Causal Discovery under Latent Confounding using Ancestral GFlowNets

·288 words
Tiago Da Silva Eliezer De Souza Da Silva Adèle Ribeiro António Góis Dominik Heider Samuel Kaski Diego Mesquita Publication Preprint Generative Flow Networks Causal Structure Learning Causal ML

We introduce a causal discovery method that estimates uncertainty and refines results with expert feedback. Using generative flow networks, we sample belief-based ancestral graphs that captures latent-confounding, and iteratively reduce uncertainty through human input, with a human-in-the-loop approach.

Prior Specification for Bayesian Matrix Factorization via Prior Predictive Matching

·353 words
Eliezer De Souza Da Silva Tomasz Kuśmierczyk Marcelo Hartmann Arto Klami Publication Journal JMLR Bayesian Matrix Factorization Theoretical Prior Predictive Analysis

A method for prior specification by optimizing hyperparameters via the prior predictive distribution. This approach matches virtual statistics generated by the prior to certain target values. We apply it to Bayesian matrix factorization models, obtaining a close-formula for the rank of the latent variables, and analytically determine the matching hyperparameters, and extend it to general models through stochastic optimization.