Metadata-Version: 2.1
Name: machine_learning_models
Version: 0.0.1
Summary: Implementations of most popular machine learning algorithms
Home-page: https://github.com/mgrzanka/MachineLearning.git
Author: Małgorzata Grzanka
Author-email: mgrzanka45@gmail.com
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Requires-Python: >=3.11.2
Description-Content-Type: text/markdown
License-File: LICENSE

# 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
```python
pip install not_existing_still_in_progress_;o
```

#### Author
Małgorzata Grzanka
