Vu Dinh (Ewing 312)
  Class times: MWF 12:20pm-1:10pm, Purnell Hall 236
  Office hours: MW 3:00pm-4:00pm, Ewing Hall 312, or by appointment.

• Syllabus

• Final project's info

• Sample report (please acknowledge the authors of the report if you use the sample).

• Reference
  • An Introduction to Statistical Learning, James, Witten, Hastie, and Tibshirani.
    The pdf of the book is available for free at http://www-bcf.usc.edu/~gareth/ISL/
  
  
  • The course provides an introduction to the fundamental techniques used in data science. The main objective of the course is to develop a good practical knowledge and a mathematical understanding of the common tools that are used to analyze modern datasets. The course also provides hands-on experience in data analysis through practical homework and class projects.

• Module 1: Introduction to supervised learning with scikit-learn and Keras
  
  • Lecture 1 (Mon, 02/07).
  • Lecture 2 (Wed, 02/09).
    Python Demo
  • Numpy Demo (Fri, 02/11).
  • Lecture 4 (Nearest neighbours) (Mon, 02/14).
    Nearest Neighbours Demo
  • Lecture 5 (Nearest neighbours (cont.)) (Wed, 02/16).
    
  • Lecture 6 (Logistic regression) (Fri, 02/18).
    Logistic regression Demo (template). (Mon, 02/21).
    Logistic regression Demo (codes).
  • Lecture 8 (Feed-forward neural networks) (Wed, 02/23).
    Feed-forward neural network Demo (template).
    Feed-forward neural network Demo (codes).
  • Lecture 9 (Feed-forward neural networks (cont.)) (Fri, 02/25).
  • Lecture 10 (SGD and Back-propagation) (Mon, 02/28).
  • Lecture 11 (Deep learning with Keras) (Wed, 03/02).
    MLP Keras Demo (template).
    Link to Goole Doc of the demo.
    MLP Keras Demo (completed).
  • Lecture 12 (Convolutional neural networks) (Fri, 03/04).
    CNN Demo (template).
    Link to Goole Doc of the demo.
    CNN Demo (completed).
  • Lecture 13 (Convolutional neural networks (cont.)) (Mon, 03/07).

• Module 2: Mathematical techniques in data science
  • Statistical learning theory
    Note: A short introduction to statistical learning theory (Posted: 03/09, Updated: 03/22)
    Lecture 14: Hypothesis spaces and loss functions (Wed, 03/09)
    Lecture 15: Hoeffding's inequality (Fri, 03/11)
    Hand-written notes: Lecture 16, Lecture 17, Lecture 18
    Lecture 19: Generalization bounds using covering number (Mon, 03/21)
  • Support vector machines
    Slides (Mon, 04/04. Updated Mon 04/11)
  • Shrinkage methods
    Slides (Mon, 04/11. Updated Tue 04/19).
    Note: Model consistency of lasso
    Further readings:
    
    • Zhao and Yu (2006).
    • Wainright (2009).
    • Sparsity, the lasso, and friends (Ryan Tibshirani).
    Lecture 29: Cross-validation
    Demo: Lasso and cross-validation
  • Boosting
    Lecture 30: Decision trees
    Demo: Decision trees
    Lecture 31: Random forests
    Lecture 32: Boosting
    Demo: Ensemble methods
    
  • Manifold learning
    Lecture 33: PCA
    Lecture 34: MDS, Isomap, LLE
    Lecture 35: t-SNE and Laplace Eigenmap
    Demo: Manifold learning
    
  • Clustering
    Lecture 36-37: Clustering
    Demo: K-means clustering
    
  
  
  
  • Assignments
    • Homework 0 (due 02/21)
      
    • Homework 1 (due 03/07)
      
    • Homework 2 (due 03/21)
      
    • Homework 3 (due 04/11)
      
    • Homework 4 (due 05/02)