Class times: Wednesday, 5:30 – 7:45pm, LL 1124
Instructor: Prof. Daniel D. Leeds (my homepage)
Teaching assistant: Amy Feng
Office: LL 610H for Office Hours; normally JMH 332 (Rose Hill)
E-mail:
Office hours: Instructor - Wed 4-5pm; TA - Wed 2-3pm (Sixth floor hallway)

Full syllabus

Course text: No book will be required, but the following will be useful as references.

• "Pattern Recognition and Machine Learning" by Bishop
• "Machine Learning: A Probabilistic Approach" by Murphy

• In Python, you can use numpy, scipy and matplotlib packages to write acceptable code for this class in Python. (You also may used the pandas package; you may not use packages like tensorflow for this class.)

1. Resources
2. Announcements
3. Slides
4. Assignments

Announcements:
Final exam is May 1, 5:30-7:30pm !
Apr 24, 12:30am: I am hoping to cover a few new topics at the beginning of class (Dirichlet processes and more notes on CNNs). Any new material today will not be tested on the final. The rest of class will be dedicated to: course evaluation (SEEQ), final project thoughts, and final exam review.
Apr 2, 4:30pm: Dr Leeds' office hours tomorrow (Apr 3) will be 3-4pm
Feb 19, 12:05am: The midterm will be Wednesday March 13.
Jan 18, 11am: Amy will have her first office hours Wed Jan 23, 2-3pm on the sixth floor of LL.

Slides: Check slides after lectures as well for updates made during class!
I have not found any one textbook or online resource to be an optimal match for the material we cover at the mathematical level we cover it. However, Andrew Ng's lecture notes, available on Stanford's Machine Learning course web site, often can be a helpful online read. I recommend these notes as well as looking through one of our course textbooks.

	Supplementary reading
Lecture 1, Course logistics, background math, intro to classifiers.
Lecture 2, Bayes classifier.	Ng notes 2 particularly pages 8-11
Lecture 3, Logistic classifier.	Ng Notes 1 Elements of Part II (starting on page 16)
Lecture 4, Support vector machines.	Ng notes 3 parts of pages 1-20
Lecture 5, Dimensionality reduction.	Ng notes 10 on PCA and Ng notes 11 on ICA
Lecture 6, Neural networks.	Chapter 1 and Chapter 2 of "Neural Networks and Deep Learning
Lecture 7, Hidden Markov Models.	Stanford notes
Lecture 8, Learning Theory.	Lecture Notes
Lecture 9, Mixture Models.
Lecture 10, Convolutional Neural Network.

Assignments:
Homework 0 - Due Jan 23 (written) and Jan 26 (code); correction uploaded Jan 18, 11am

Additional corrections added to A7c and B4,5,6 (log function) Jan 22, 7:25pm. You may submit your answer based on the original question OR based on the corrected question.

Homework 0 answers

Code grade:

27-30 A range

24-27 B range

21-24 C range

Written grade:

74-83 A range

64-74 B range

55-64 C range

Additional correction to Question B8 (labelMP2), uploaded Feb 4 at 11pm.

HW 1 answers

Code grade:

36-41 A range

30-36 B range

22-30 C range

Written grade:

65-73 A range

57-65 B range

49-57 C range

Feb 21: I added a note to C4 - all λs must be non-negative. You can force negative λs to 0 at each learning step.

Questions A6b corrected and C5 clarified Feb 23rd.

Clarification to A5 and C1 and C5 added Feb 26th.

In addition to the numpy functions mentioned in the numpy python pointers, I also will allow np.concatenate and np.absolute for this homework. You cannot assume any other numpy functions are allowed if they were not explicitly named in the homework or the numpy pointer slides (except in Q5).

Note, the matrix "data" was written intended for Q5; the matrix "data3" was written intended for Q2. For Q5, I will accept your answer on either data or data3 or both.

HW2 written answers

72-82 A range

60-72 B range

50-60 C range

HW2 Code Grade

32-37 A range

27-32 B range

22-27 C range

HW3 written answers

Midterm answers

77-87 A range

65-77 B range - changed curve

53-65 C range - changed curve

HW3 Code Grade

39.5-44 A range

35-39.5 B range

30.5-35 C range

Fake, Ungraded HMM Homework - due never
HMM HW Answers