MLearn

A collection of machine learning algorithms, written in Python.

Supported Algorithms

Supervised Learning

• Decision Trees [in progress]
• Linear Regression
  • L1, L2 Regularization
  • Elastic Net Regularization [in progress]
  • Normal Equation Method
• Logistic Regression
  • L1, L2 Regularization
• K-Nearest Neighbors (KNN): regressor, classifier
  • Distance-weighted KNN [in progress]
• Multilayer Perceptron [planned]
• Naive-Bayes methods
  • Gaussian NB
• Support Vector Machines [in progress]
• Ensembles of learners [planned]

Unsupervised Learning

• K-Means Clustering [planned]

Optimizers

• Gradient descent and variants
  • Batch gradient descent
  • Stochastic (mini-batch) gradient descent [in progress]
• Adam
• RMSProp
• Momentum

Data Preprocessing Tools

• Feature normalization/scaling
• Nonlinear feature generation

Usage

Import the desired learning algorithm from algorithms. Then, use the .fit(X, y) method to train the algorithm, and .predict(X) to make predictions.