ML-Algorithm

_ _

Introduction

Implement machine learning algorithms with python without sklearn. Some classes are design especially for CQU-ML course by Professor.He

Install

• With pip : pip3 install cquai-ml
• With src : Clone or fork this project, then build it with python3 setup.py install

Usage

In most cases, API in this project is similar to scikit-learn project.

For example, if you want to run a decision tree classifier based on C4.5 (While scikit-learn use opt-CART instead of C4.5)

from cquai_ml import DecisionTreeClassifier
from sklearn.datasets import load_breast_cancer # get a dataset

X, y = load_breast_cancer(return_X_y=True)

clf1 = DecisionTreeClassifier(max_depth=1).fit(X, y)
pred1 = clf1.predict(X)

Contributing

Everyone is welcomed to contribute!

Note: To maintain code cleanliness, besides functional, this project should also be code-quality-issue-free. In other words, code should pass Bandit, Prospector and PyLint with no warning. Click the code quality badge above for more information.

We currently provides:

• DatasetSpace
• UnionHypothesisSpace
• LinearRegression
• LogisticRegression
• LinearDiscriminantAnalysis
• KNeighborsClassifier
• DecisionTreeRegressor
• DecisionTreeClassifier