ENEM-2016-ML-EXERCISES

This repository contains jupyter notebooks with the analysis and application of regression and classification techniques to the ENEM 2016 dataset, which contains students' informations as well as their grades. It's based in an exercise from the Aceleradev Data Science program by Codenation.
In the ENEM_2016_regression.ipynb notebook, you'll find the clean up and analysis process of the dataset, as well as the application of linear regression, ridge regression and lasso techniques.
In the ENEM_2016_classification.ipynb notebook, you'll find the clean up and analysis process of the dataset, as well as the application of logistic regression with and without cross validation, and random forests classifier.
The train and test sets used are also available in the csv files train.csv, test_regression.csv and test_classification.csv. The test datasets do not contain the answers though, as the models' predictions/classifications were to be evaluated by a third party website. You can split the train dataset into train and test.
The data dictionary is available as it were distributed in the OpenDocument spreadsheet Dicionario_Microdados_Enem_2016. It is in portuguese because ENEM is a brazilian national test to evaluate high school students, and it's a mean to access many brazilian universities.
A requirements.txt is also available. Have fun!