Machine Learning

Supervised Learning

You have the output column and you train the model with it.

• Classification: Predict a class label or category (e.g., true/false, spam/ham, cat/dog).

  • Logistic Regression (binary or multi-class)
  • Decision Tree
  • Random Forest
  • Support Vector Machine (SVM)
  • K-Nearest Neighbors (KNN)
  • Naive Bayes
  • Gradient Boosting Machines (GBM)
  • Neural Networks

• Regression: Predict a continuous value (e.g., price, temperature, age).

  • Linear Regression
  • Decision Tree Regression
  • Random Forest Regression
  • Support Vector Regression (SVR)
  • K-Nearest Neighbors Regression
  • Ridge Regression
  • Lasso Regression
  • Polynomial Regression
  • Neural Networks (e.g., Multilayer Perceptron)

Unsupervised Learning

You don't have the output column and you train the model without it.

• Dimensionality Reduction: Reduce the number of features while preserving important information.

  • Principal Component Analysis (PCA)
  • t-Distributed Stochastic Neighbor Embedding (t-SNE)
  • Singular Value Decomposition (SVD)
  • Linear Discriminant Analysis (LDA)
  • Independent Component Analysis (ICA)

• Density Estimation: Estimate the probability distribution of data.

  • Gaussian Mixture Models (GMM)
  • Kernel Density Estimation (KDE)

• Market Basket Analysis: Identify associations between items.

  • Apriori Algorithm
  • Eclat Algorithm

• Clustering: Group similar data points together.

  • K-Means
  • Hierarchical Clustering
  • DBSCAN (Density-Based Spatial Clustering of Applications with Noise)
  • Mean Shift

Metrics

Summary:

• Accuracy gives an overall measure of correct predictions.
• Confusion Matrix breaks down the types of correct and incorrect predictions.
• Recall (or sensitivity) focuses on correctly identifying positive cases.
• Specificity (or true negative rate) focuses on correctly identifying negative cases.
• F1 Score balances precision and recall into a single metric, useful when there's an uneven class distribution.