Pennsylvania State University

Course Syllabus

• Overview of scientific and practical rationale, applications
• Statistical Machine Learning Theory and Applications
  1. Decision-theoretic foundations of classification
  2. Maximum likelihood, maximum a posteriori and Bayesian Frameworks
  3. Probabilistic generative models (with emphasis on the exponential family)
  4. Probabilistic discriminative models
  5. Kernel machines
  6. Representative algorithms
• Algorithmic Learning Theory and Applications
  1. Algorithmic models of learning
  2. Mistake bound model – winnow and multiplicative update algorithms
  3. PAC Model – sample complexity, algorithmic complexity, PAC-easy and PAC-hard learning problems
  4. Occam algorithms
  5. Boosting and ensemble learning
  6. Curriculum learning - Learning under helpful distributions
• Probabilistic graphical models
  1. Bayesian networks - structure and parameter learning
  2. Markov models and Hidden Markov Models, and their generalizations
  3. Latent variable models - topic models
  4. Probabilistic relational models
  5. Stochastic process models
  6. Grammars
• Selected Topics in Advanced Machine Learning
  1. Multi-instance learning
  2. Multi-label learning
  3. Multi-relational learning
  4. Structured label learning
  5. Distributional learning
  6. Deep learning
  7. Learning from longitudinal data
• Learning Predictive Models from Big Data
  1. Scaling up learning algorithms to large data sets
  2. Scaling up learning algorithms to high dimensional data sets - feature selection, abstraction, hashing
  3. Learning predictive models from distributed data
  4. Learning predictive models from linked data
  5. Big Data Analytics Platforms and Tools