Skip to content

Latest commit

 

History

History
35 lines (18 loc) · 1.29 KB

File metadata and controls

35 lines (18 loc) · 1.29 KB

Multiclassifier

This is a project is an implemention of a multiclassifier, using the one-vs-all (or one-vs-rest) technique, which aim to predict the Harry Potter's Hogwarts Houses of a student according to a set of features.

Usage

describes.py dataset display mean, standard deviation, minimun, maximum, count, median, first and third quartile for all features.

histogram.py dataset show the distribution of grades for all features between all the houses.

scatter_plot.py dataset display the a scatter plot of the two houses with the strongest relation according to the Bravais-Pearson relation.

pair_plot.py dataset show a pair plot of all numeric values of the dataset.

logreg_train.py dataset train the model with the dataset and generate a weight.py file with all the thetas needed for prediction.

logreg_predict.py dataset weights generate prediction for the given dataset and weights. It generate a houses.csv file with all the result of the prediction.

Images

Describes

describes.py screenshot

Histogram

histogram.py screenshot

Scatter Plot

scatter_plot.py screenshot

Logistic Regression

logistic regression screenshot