About me

I work as a Research Assistant focused on Machine Learning at University of Vale do Rio dos Sinos in São Leopoldo, Brazil. I hold a Bachelor’s degree in Computer Science from the same university, with an Outstanding Student Award from the Brazilian Computing Society (SBC). My main research interests are centered around Reinforcement Learning, more specifically with Options and dynamic Option creation. I have also worked on projects with extractive NLP, LLMs, and Neural Networks.

On my spare time, I use my programming skills to create and contribute to open-source projects, mostly related to the Linux desktop. Most of my OSS efforts go towards Vanilla OS, where I act as Contributors Leader. My contributions to Vanilla ranges from creating graphical applications with GTK to low-level programs using Go. I also make occasional contributions to projects from the GNOME Desktop Environment, such as GNOME Builder and GtkSourceView.

Programming Language Experience

• Go:
  • Low-level tools like ABRoot and Albius
  • REST APIs like Differ
  • CLI tools like Apx
• Python:
  • ML code with Pytorch and Numpy
  • GTK applications with PyGOBject like Vanilla Installer
• Lua:
  • Neovim plugins like JABS.nvim
• Rust:
  • ML code using Ndarray, awaiting publication
• Bash:
  • General Linux automation
• C:
  • Both GTK applications and Linux native programs
• Zig:
  • Mostly experimental projects like spreadr

---

Publications

• M. B. Melchiades, and G. de O. Ramos, “Dynamic Option Creation for Option-Critic Reinforcement Learning”. (Undergraduate Thesis)
  • PDF
• M. B. Melchiades, C. D. P. Crovato, E. Nedel, and L. V. Schreiber, “FastIoT: an efficient and very fast compression model for displaying a huge volume of IoT data in web environments”.
  • Website
• A. de Souza, M. B. Melchiades, S. J. Rigo, and G. de O. Ramos, “MoStress: a Sequence Model for Stress Classification,” in 2022 International Joint Conference on Neural Networks (IJCNN), Padua, Italy: IEEE, Jul. 2022, pp. 1–8.
  • Website
• M. Melchiades, L. Schreiber, G. Ramos, C. Paredes, R. Goytia, and R. Da Rosa, “Anticipating faults by predicting non-linearity of environment variables with neural networks: a case study in semiconductor manufacturing,” in LatinX in AI at International Conference on Machine Learning 2021, Journal of LatinX in AI Research, Jul. 2021.
  • PDF