Basics of AI - implementations of ML models and optimization methods

This project is an implementation of the logistic regression model as well as optimization and regularization methods/algorithms using python and numpy libraries, with practical examples and explanations to better present the concepts.

All results of our models are going to be compared with the ready implementations found in the sklearn library.

Introduction

Logistic regression, despite its name, is a linear model for classification rather than regression. Logistic regression is also often called logit regression because it uses a logistic function to predict the results. LR is a probabilistic model that assigns the probability of a sample to a given class. You can find more about linear regression here.

Interface

Our models provide a clean interface based on the sklearn library interface. In addition to the models themselves, various regularizers and optimization algorithms are available.

More details can be found inside the docs.

Basic Example

model = LogisticRegression()

model.fit(X_train, y_train)
y_pred = model.predict(X_test)
score = model.evaluate(y_test, y_pred)

You examples can be found inside the examples section 🔥.

Credits

• images: https://www.equiskill.com/