This repository contains other projects I have developed and were important for my Machine Learning journey. I have also included some projects that I have developed for classes, but that I believe are relevant to my portfolio.
They are organized by when I finished their development. I have also included a brief description of the project and the main subjects I have improved upon while developing them.
Finally, the main subjects can be roughly divided into Machine Learning and Optimization Projects, and I believe that it is by understanding both and how they can be used that we can develop the best solutions for the problems we face.
For my application to Digital Product School (DPS), the AI track received the challenge of predicting the total number of accidents (under accident_type) in Munich for the year of 2020.
The main subjects I have improved upon were:
- Time Series Analysis
- Deployment of Machine Learning Models
You can access it here!
This project was developed for Pontifical Catholic University of Rio de Janeiro's master level Neural Networks class. It develops a post processing pipeline for precipitation numerical forecast for an area in Brazil, coordinates (-23.0, -54.0) and (-16.0, -46.0).
I believe this is the most ambitious project I have developed yet, specially because it was developed in a short period of time. The main subjects I have improved upon were:
- Neural Networks
- Bibliographic Research (on weather)
You can access it here!
This repository is the result of a semester at a PUC-Rio's course by the same name of the book Machine Learning Under a Modern Optimization Lens.
The main subjects I have improved upon were:
- Optimization
- Combining Optimization and Machine Learning
You can access it here!
This project was the result of a short two week project for a selective process. The challenge was to make SKU predictions for a Brazilian retailer.
The main subjects I have improved upon were:
- Time Series Analysis
- Communication of Results
- Data Storytelling
You can access it here!
This project was developed for Pontifical Catholic University of Rio de Janeiro's master level Time Series class. It compares different models - SARIMAX, VAR and Neural Networks - for wind speed forecasting.
The main subjects I have improved upon were:
- Time Series Analysis
- Technical Communication
You can access it here!
This project was developed for Pontifical Catholic University of Rio de Janeiro's master level Decision and Risk Analysis class. It is a mathematical optimization model for the Maximum Coverage with Minimum Cost problem.
It has been published in the Brazilian Operations Research Simposium (SBPO) under Cobertura Máxima e Custo Mínimo para Alocação de Recursos no Tempo (pt).
The main subjects I have improved upon were:
- Optimization
- Geo visualization
You can access it here!
I hope you enjoy my portfolio and I am looking forward to hearing from you!