Projects

Machine/Statistical Learning

• Credit Card Fraud Detection: Kaggle notebook where I explore the SMOTE balancing procedure and skewness reduction technique for credit fraud detection in a highly imbalanced dataset.
  Tech: Python, pandas, numpy, seaborn, scikit-learn, imblearn, scipy. [link]
• GA-evolving: Undergrad project on evolutionary computation algorithms. I’ve implemented a generic genetic algorithm (GA) in python along with some common genetic operators.
  Tech: Python, pandas, numpy, seaborn, matplotlib. [link]
• Time-series LDA: Used the latent Dirichlet allocation (LDA) algorithm to cluster a fuzzy time-series. The ideia is to create a word representation of a time-series and use LDA to detect change points in it. Tech: MATLAB, R. [link]
• GA-rp: Undergrad project on nonlinear optimization using genetic algorithm (GA). I’ve implemented a GA with real representation using a polarized crossover operator.
  Tech: MATLAB. [link]
• N-Puzzle Solving the N-puzzle problem with two informed search strategies: Best First and A*.
  Tech: C, C++. [link]
• N-Queens Solving the N-queens problem using the Hill-Climbing method. Tech: C. [link]

Web Development/Data Engineering

• CKL-Challenge: I created a web app that provides an API according to the REST architecture. The API periodically scrapes news from four different sources (twitter statuses, google plus feeds, RSS system, HTML parsing) and stores them into a PostgreSQL database. The system has continuous integration and deployment.
  Tech: Python, Django, Django-rest-framework, RESTFul, CI/CD, Celery, PostgreSQL, CodeShip, Heroku, AWS SQS. [link]
• Reviewer-API: I created an API for a simple company review system that allows users to create/update/delete/retrieve their reviews.
  Tech: Python, Django, Django-rest-framework, RESTFul, CI/CD, PostgreSQL, Travis, Docker, Sentry, Heroku. [link]

Desktop Apps

• ColorMixer: undergrad project of a supervisor system to explore different aspects of a desktop application, such as, login, splashscreen, log system, database connection, internationalization, etc. Tech: C++, wxWidgets, SQLite. [link]

General

• Compression: undergrad project to implement three compression algorithms: LZ78, Huffman Tree and Bit Map.
  Tech: C++. [link]
• PyTvName: a small python project to explore regular expression in data transformation. The ideia is extracting structured information, such as, episode number, season, title, and show name, from a generic TV show episode file name.
  Tech: Python, regex. [link]

Open Source Contributions

• PYFTS/pyFTS (Python): Added the dynamic subtractive clustering technique proposed by Chiu in 1994.
• dleitee/strman (JS): Fixed the optional parameters of a JS function that tested strings ending with some value.
• keon/algorithms (Python): Added the Tarjan’s strongly connected components finder algorithm.
• sagivo/algorithms (Ruby): Added the Tarjan’s strongly connected components finder algorithm.
• FrankKair/polyglot-euler Added solutions to Project Euler problems using Python, Ruby and C.