Machine Learning From Scratch (mlfs)

A repository containing ML algorithms implemented from scratch (python + Numpy)

The repo currently has vectorized implementations + examples for the following algorithms:

ML algorithms:

Supervised Learning:

• One variable linear regression
• Multi-variable linear regression
• Multi-variable logistic regression
• Decision Tree Classifier
• Decision Tree Regressor
• Random Forest Classifier
• Random Forest Regressor

Unsupervised Learning:

• K-means clustering
• Anomaly detection using multivariate normal distribution
• PCA
• Recommender systems
  • Collaborative Filtering (TF2 implementation)
  • Content-based Filtering (TF2 implementation)

DL algorithms:

Supervised Learning:

• Neural Network

• Dense Layer

• Optimizers:

  • SGD (stochastic gradient descent)

• Activations:

  • Sigmoid
  • Linear
  • ReLU

• Losses:

  • Mean Squared Error
  • Categorical Crossentropy
  • Categorical Crossentropy from logits