Portfolio

Data Science

PIMA Indian Diabetes Prediction

The classic PIMA Indian Diabetes data set, but with a more in-depth approach. In this analysis, my goal is to explore the unbalanced binary class and find a method that makes my prediction more assertive. I compare some models with the original data to Undersampling Majority Class, Oversampling Minority Class, SMOTE and SMOTE combined with Undersampling.

E-Commerce Data Analysis

This work is based on a data set of transnational sales from an UK-based non-store online retail. In this analysis, my goal is to explore the data, learn more about sales and find patterns.

Houses for Rent

This data set is based on Brazilian houses for rent. It is a small project and the objective is to explore data, understand the relationship between features and rent and train models for rent forecasting.