2.7. Data Science and Machine Learning Mathematics (DSM)

2.7. Data Science and Machine Learning Mathematics (DSM)

This area provides the mathematical foundations underlying modern data science and machine learning. It covers statistical learning theory, supervised and unsupervised learning methods, model evaluation techniques, and the mathematical principles that govern the behavior of learning algorithms. The focus is on understanding why machine learning methods work, their theoretical guarantees, and their limitations, rather than just their application.

Table 2.7: List of KUs in the Data Science and Machine Learning Mathematics area.

2.7.1. DSM/Foundations of Statistical Learning  (Core Tier1: 1 hr, Core Tier2: 1 hr) ↑ Back to top

Topics:
Core

  • Bias-variance tradeoff
  • Vapnik-Chervonenkis (VC) theory
  • Empirical risk minimization
  • Generalization bounds
  • Learning theory fundamentals

Learning Outcomes:
Core:

  1. Analyze the bias-variance tradeoff in model selection [Assessment]
  2. Apply VC theory to analyze model complexity [Usage]
  3. Implement empirical risk minimization [Usage]
  4. Interpret generalization bounds [Familiarity]

2.7.2. DSM/Supervised Learning Methods  (Core Tier1: 1 hr, Core Tier2: 1 hr) ↑ Back to top

Topics:
Core

  • Linear models and extensions
  • Support Vector Machines (SVM)
  • Decision trees and ensemble methods
  • Kernel methods
  • Probabilistic supervised learning

Learning Outcomes:
Core:

  1. Implement and analyze linear models [Usage]
  2. Apply SVM for classification and regression [Usage]
  3. Construct ensemble methods (bagging, boosting) [Usage]
  4. Use kernel methods for nonlinear problems [Usage]

2.7.3. DSM/Unsupervised Learning Methods  (Core Tier1: 1 hr, Core Tier2: 1 hr) ↑ Back to top

Topics:
Core

  • Clustering algorithms (k-means, hierarchical)
  • Dimensionality reduction (PCA, t-SNE)
  • Density estimation
  • Manifold learning
  • Autoencoders and deep unsupervised learning

Learning Outcomes:
Core:

  1. Apply clustering algorithms to data exploration [Usage]
  2. Implement PCA and other dimensionality reduction methods [Usage]
  3. Use density estimation for unsupervised learning [Usage]
  4. Analyze manifold learning techniques [Assessment]

2.7.4. DSM/Model Evaluation and Selection  (Core Tier1: 1 hr, Core Tier2: 1 hr) ↑ Back to top

Topics:
Core

  • Cross-validation strategies
  • Performance metrics for classification and regression
  • Model comparison and statistical tests
  • Hyperparameter optimization
  • Learning curves and diagnostic tools

Learning Outcomes:
Core:

  1. Implement cross-validation for model assessment [Usage]
  2. Select appropriate performance metrics [Assessment]
  3. Compare models using statistical tests [Usage]
  4. Optimize hyperparameters using systematic methods [Usage]

Spotted a typo, an outdated course, a broken link, or have a suggestion? Let us know.

Scan to open on your phone