Personal projects

São Paulo State Hospital Registry of Cancer Database

A short joint project developed at Information Technology Center Renato Archer. Performed exploratory data analysis on the Hospital Registry of Cancer from the state of Sao Paulo, Brazil. It will include soon a data cleaning section intended to make the dataset more machine learnable.

Predicting patient survival rates

At first, an exploratory data analysis was performed. Then several baseline models were built, and the best ranked models by the F1 score were selected and improved since this was the evaluation criteria for this task during my participation in the DPhi Bootcamp.

Predicting employee retention using logistic regression

The LogisticRegression() classifier was used to build a model to predict wheather or not an employee will leave the job based on several features including salary, level of satisfaction, etc.

Predicting crime rate with decision trees

Four predictive models for crime rate were built using simple decision trees and their performances were measured by analyzing the respective learning curves. The complexity of the model was also reduced by tuning some regressor parameters.

Probing eBay Kleinanzeigen used car ads

This project focus on a reduced dataset of used cars from eBay Kleinanzeigen, a classifieds section of the German eBay. After cleaning the data, the used car listings are analyzed.

Exploring trends in the dataset of Hacker News site

Hacker News (HN) is a social news website that mixes computer science and entrepreneurship. This project explores some trends in the data set of HN site, more specifically, it focus on submissions where users ask questions (Ask HN posts) or show a project, product or something interesting (Show HN posts).

Mapping apps for profitability at the App Store and Play Store

This is about understanding data and using the knowledge acquired in this process to improve business strategy. The two data sets used here are from App Store and Play Store. It focus on free English apps, trying to understand what type of apps would probably attract more users. At the end, three new app promising ideas are suggested.

What I learned when trying to transcribe an audio file with Python

Once a month I attend a meeting at my Department when audio is recorded to make the writing of the minutes easier. This simple project is an attempt to automate the transcription of this audio and to make it one of my adventures using JupyterLab and Python.