machine-learning-models 0.0.1

Implementations of most popular machine learning algorithms

About this repo

This repository contains my own implementations of the most popular machine learning models. It also has scripts for data preproccessing (datasets titanic and GaltonFamilies). For classification, I implemented my own class for nested cross validation. There are functions to evaluate model using confusion matrix (calculating accurancy, sensitivity, specificity, precision etc.) as well as to draw ROC curve.

Note

All of the models and validation were implemented by myslef, without using sklearn library. They were done for learning purposes. There is a seperate file named sklearn.py, in which I focused on exploring sklearn library

Models in this repo

1. Linear models

• Linear Regression
• Logistic Regression

2. Trees

• Regression Tree
• Decision Tree
• Random Forest

3. Boosting

• Adaboost
• Gradient boosting
• Extreme Gradient Boosting

4. SVM

• SVM primal bez jądra
• SVM dual z jądrem

5. Naive Bayes
6. K-nearest-neighbours

Python package

This repo can be installed as a package via command

pip install not_existing_still_in_progress_;o

Author

Małgorzata Grzanka