Classical Machine Learning Algorithms

This repository contains the machine learning algorithms implemented in Python 3.5 from scratch :

1. K-Nearest Neighbors
2. K-Mean Clustering
3. Regression Techniques
   • Linear Regression (Univariate & Multivariate)
   • Logistic Regression (Binary Class and Multi Class)
4. Decision Trees (CART using GINI impurity)
5. Support Vector Machines
   • Linear SVM
   • Multi-Class SVM
6. Naive Bayes (Gaussian)
7. Two Layer Neural Network
8. Similarities / Distance
   • Jaccard Similarity
   • Cosine Similarity
   • Centered Cosine Similarity / Pearson Correlation
9. Recommender Systems
   • Content Based
     • Using Linear Regression
   • Collaborative Filtering
     • Memory Based using Pearson Similarity
     • Model based using Linear Regression
10. Dimensionality Reduction
    • Feature Selection
      • Filter Methods
        • Select K Best using Chi Squared Test
        • Variance Threshold
        • Information Gain
      • Wrapper Methods
        • Recursive Feature Elimination (RFE)
    • Feature Extraction
      • Principal Component Analysis (PCA)
      • Auto Encoder
      • t-SNE

Steps for running

1. virtualenv .Classical-ML
2. source .Classical-ML/bin/activate
3. pip3 install -r requirements.txt
4. python3 main.py