This project focuses on implementing a few fundamental Machine Learning (ML) algorithms, including Categorical Naive Bayes, mini-batch logistic regression, and the One vs Rest approach. The objective is to build these algorithms from scratch, along with essential preprocessing tools like train_test_split, One Hot encoding, and Standard scaler, using only Python's numpy library. These implementations will be applied to classify datasets such as iris, breast_cancer, and credit.
The file clasificadores.py contains the machine learning implementations from scratch and their use in some dataset. To run it use python clasificadores.py.
Only need to install numpy, scikit-learn and scipy.