Daniel Augusto

Hi! I am a final-year Ph.D. student at UCL, advised by Prof. Marc Deisenroth, with a focus on probabilistic machine learning and Gaussian processes. I am particularly interested in their applications to climate modeling and material discovery.

Beyond research, I am passionate about software engineering, emerging technologies, and translating academic insights into deployable solutions, which has led me to multiple roles in industrial research, DevOps, and software engineering.

This webpage contains my contact information and a list of publications. If you are interested my resumé can be downloaded here.

Bibliography

2024
- Variational Inference with Censored Gaussian Process Regressors.
  Andrea Karlova, Rishabh Kabda, Daniel Augusto de Souza, and Brooks Paige.
  In: ICML 2024 Workshop on Structured Probabilistic Inference & Generative Modeling
- Steaming Bayes GFlowNets.
  Tiago da Silva, Daniel Augusto de Souza, and Diego Mesquita.
  In: Advances in Neural Information Processing Systems (NeurIPS) 37
2023
- Actually Sparse Variational Gaussian Processes.
  Harry Jake Cunningham, Daniel Augusto de Souza, So Takao, Mark van der Wilk, and Marc Peter Deisenroth.
  In: 26th International Conference on Artificial Intelligence and Statistics (AISTATS)
- Thin and Deep Gaussian Processes.
  Daniel Augusto de Souza, Alexander Nikitin, S. T. John, Magnus Ross, Mauricio A. Álvarez, Marc Peter Deisenroth, João P. P. Gomes, Diego Mesquita, and César Lincoln Mattos.
  In: Advances in Neural Information Processing Systems (NeurIPS) 36
  
  Blogpost
2022
- Parallel MCMC Without Embarrassing Failures.
  Daniel Augusto de Souza, Diego Mesquita, Samuel Kaski, and Luigi Acerbi.
  In: 25th International Conference on Artificial Intelligence and Statistics (AISTATS)
- Deep Mahalanobis Gaussian Process.
  Daniel Augusto de Souza, Diego Mesquita, César Lincoln Mattos, and João Paulo Gomes.
  In: NeurIPS Workshop on Gaussian Processes, Spatiotemporal Modeling, and Decision-making Systems
2021
- Learning GPLVM with arbitrary kernels using the unscented transformation.
  Daniel Augusto de Souza, Diego Mesquita, João Paulo Gomes, and César Lincoln C. Mattos.
  In: 24th International Conference on Artificial Intelligence and Statistics (AISTATS)
2019
- No-PASt-BO: Normalized Portfolio Allocation Strategy for Bayesian Optimization.
  Thiago de Paula Vasconcelos, Daniel A. R. M. A. de Souza, César Lincoln C. Mattos, and João Paulo P. Gomes.
  In: 2019 IEEE 31st International Conference on Tools with Artificial Intelligence (ICTAI)
- Evaluation of Data Based Normal Behavior Models for Fault Detection in Wind Turbines.
  Daniel Souza, Thiago Vasconcelos, César Mattos, and João Gomes.
  In: 2019 8th Brazilian Conference on Intelligent Systems (BRACIS)

Thesis

Contributions on latent projections for Gaussian process modeling.
Daniel Augusto Ramos Macedo Antunes de Souza.
M.Sc. thesis. Jointly supervised by: João Paulo Gomes, César Lincoln C. Mattos. Universidade Federal do Ceará

Bibliography

2024

2023

2022

2021

2019

Thesis