ML-Mono

Collections of rough implementations of various ML concepts for learning purposes.

Project Structure

This repository is organized into several directories, each focusing on different aspects of machine learning.

Supervised Learning

Implementations of supervised learning algorithms:

• Linear Regression: See supervised/linear_regression.py.
• Logistic Regression: See supervised/logistic_regression.py.
• Neural Networks:
  • Recurrent Neural Networks (RNNs):
    • Simple RNN: supervised/neural_networks/rnn/simple_rnn.py.
    • LSTM (Long Short-Term Memory): supervised/neural_networks/rnn/lstm.py.
    • RNN-based Language Model: supervised/neural_networks/rnn/language_model.py.

Unsupervised Learning

Implementations of unsupervised learning algorithms:

• Clustering:
  • K-Means Clustering: unsupervised/clustering/kmeans.py.

Metrics

Currently implemented metrics for model evaluation:

• Accuracy: For evaluating classification models. See metrics/classification_metrics.py.

Notebooks

Jupyter notebooks demonstrating the implementation and usage of various algorithms:

• Linear Regression Example: notebooks/LinRegression_Example.ipynb.
• Logistic Regression Example: notebooks/LogRegression_Example.ipynb.
• Training RNNs to Generate Content Inspired by Shakespeare: notebooks/RNN_Shakespeare_Training.ipynb.

TODO:

1. Add mixin for Unsupervised & Supervised Models to implement fit_predict, fit, predict methods + X, y fields.

License

This project is licensed under the MIT License.