1 Bayesian methods for machine learning

  • Instructor: Lizhen Lin (University of Notre Dame, Mỹ)
  • Time: Thứ năm, Thứ sáu 5–6/7, 2 buổi sáng, 9:00-11:30, Phòng F207.
  • Abstract:
    1. Parametric Bayesian analysis in Machine learning
      • Basics on Bayesian inference;
      • An intro to Markov Chain Monte Carlo (e.g., Gibbs sampler);
      • Factor analysis and dictionary learning;
    2. Non-parametric Bayesian analysis in Machine learning
      • Dirichlet process (DP);
      • DP mixture models;
      • Gaussian process for regression/classification/density estimation; Latent gaussian process models;
      • Bayesian optimizations;
    3. Neural networks and Bayesian deep learning (if time permits);

2 An Introduction to the mathematics of machine learning

  • Instructor: Hien Tran (North Carolina State University Raleigh, Mỹ)
  • Time: 16–18/7, 3 buổi, Phòng F207
  • Monday, July 16
    Lecture:     9:00-10:15
    Break:    10:15-10:45
    Lecture:    10:45-12:00

    Lecture:    2:00-3:15
    Break:    3:15-3:45
    Lecture:    3:45-5:00

    Wednesday, July 18
    Lecture:     9:00-10:15
    Break:    10:15-10:45
    Lecture:    10:45-12:00
     
  • Abstract:
    • Monday’s morning: Supervised Learning
      • Linear discriminant analysis (LDA),
      • Support vector machines (SVMs),
      • K-nearest neighbors (k-NN),
      • Classification trees (CT)
    • Monday’s afternoon: Model Evaluation and Feature Selection
      • Model evaluation: Confusion matrix, loss function, hypothesis testing,
      • Feature selection: Principal component analysis (PCA), ROC, hypothesis testing
    • Wednesday’s morning:
      • Unsupervised learning: k-means clustering,
      • Neural networks and deep learning

3 Level set method and mean curvature flow equation

  • Instructor: Hung Tran (University of Wisconsin Madison, Mỹ).
  • Abstract: I will present some basic results on the level set method and mean curvature flow equation (MCF). In particular, I will prove well-posedness of viscosity solutions to MCF. Some background on viscosity solutions can be found in Appendix of the lecture notes of Mitake and I http://www.math.wisc.edu/~hung/Mitake-Tran-LN.pdf, but are not really required to take the class.
  • Time: 16–19/7, 4 buổi, Phòng F207.