COL774: Machine Learning

Course Content

Additional Resources

• Notes on Avoiding Overfitting in Decision Trees
• On Convergence of Stochastic Gradient Descent (Youtube Video by Andrew Ng)
• On Discriminative vs. Generative Classifiers: A comparison of logistic regression and naive Bayes. Andrew Y. Ng and Michael Jordan. NIPS 2001.
• Generative and Discrminative Classifiers: Naive Bayes and Logistic Regression. Chapter 1, Machine Learning. Tom Mitchell.
• Deep Learning Book
  

Review Material

Topic	Notes
Probability	prob.pdf
Linear Algebra	linalg.pdf
Gaussian Distribution	gaussians.pdf
Convex Optimization (1)	convex-1.pdf

References

1. Machine Learning: A Probabilistic Perspective. Kevin Murphy. MIT Press, 2012.
2. Pattern Recognition and Machine Learning. Christopher Bishop. First Edition, Springer, 2006.
3. Pattern Classification. Richard Duda, Peter Hart and David Stock. Second Edition, Wiley-Interscience, 2000.
4. Machine Learning. Tom Mitchell. First Edition, McGraw-Hill, 1997.

1. You are free to discuss the problems with other students in the class. You should include the names of the people you had a significant discussion with in your submission.
2. All your solutions should be produced independently without referring to any discussion notes or the code someone else would have written.
3. All the programming should be done in MATLAB (unless otherwise allowed explicitly). Include comments for readability.
4. Code should be submitted using Moodle Page.
5. Create a separate directory for each of the questions named by the question number. For instance, for question 1, all your submissions files should be put in the directory named Q1 (and so on for other questions). Put all the Question sub-directories in a single top level directory. This directory should be named as "yourentrynumber_firstname_lastname". For example, if your entry number is "2013anz7535" and your name is "Nilesh Pathak", your submission directory should be named as "2013anz7535_nilesh_pathak". You should zip your directory and name the resulting file as "yourentrynumber_firstname_lastname.zip" e.g. in the above example it will be "2013anz7535_nilesh_pathak.zip". This single zip file should be submitted online.
6. Honor Code: Any cases of copying will be awarded a zero on the assignment and a penalty of -10. More severe penalties may follow.
7. Late Policy: You are allowed a total of 5 late days 6 late days acorss all the assignments. You are free to decide how you would like to use them. You will get a zero on an assignment once you exceed the (total) allowed exemption of 5 days 6 days.

• Pratice Questions for Major (You should also revise Practice Sets 1 and 2 below).
• Pratice Questions for Minor 2
• Pratice Questions for Minor 1

1. Assignment 4 [Updated: 11:40 pm, April 13] . Due Date: Friday April 21, 2017. 11:50 pm.
   Datasets:
   • Question 1: kmeans_data.zip
   • Question 2:
     • Dataset 1: lfw_funneled.tgz Use the updated version posted on Piazza
     • Dataset 2:att_faces.tar.Z
   • Assignment 3. Due Date: Tuesday April 4, 2017. 11:50 pm.
     Datasets:
     • Question 1: dtree_data.zip.
     • Question 2: neural_data.zip.
   • Assignment 2. Due Date: Friday March 10, 2017. 11:50 pm.
     Datasets:
     • Question 1: nb_data.zip. Stopword Removal/Stemming Script: scripts.zip.
     • Question 2: svm_data.zip.
   • Assignment 1. New Due Date: Friday February 10, 2017. 11:50 pm.
     Datasets:
     • Question 1: q1x.dat, q1y.dat
     • Question 2: q2x.dat, q2y.dat
     • Question 3: q3x.dat, q3y.dat
     • Question 4: q4x.dat, q4y.dat

Grading Policy (Tentative)