

CSC321 Spring 2014
Introduction to Neural Networks and Machine Learning
University of Toronto Mississauga
Look here at least once a
week for news about the course.
Announcements:
 Online course evaluations can be filled out until
April 9 at midnight.
 The final exam will cover support vector machines,
Boltzmann machines, restricted Boltzmann machines,
neural nets, autoencoders, Bayesian learning,
clustering, mixtures of Gaussians, and combining
models.
 The final exam will use the "I don't know" policy.
i.e., If you leave a question blank and simply write
"I don't know", you will receive 20% of the value of
the question.
 The final exam will be closed book, but you will be
allowed one page of notes (doublesided, 8.5x11 inch)
in 12point font (or larger) and no more than 12,000
characters total. No other aids are allowed.
 Don't forget to fill out your online course
evaluations.
 Another minor bug in Assignment 4 and in the code
has been corrected. Please reload both.
(March 26, 11:15am)
 A minor bug in Assignment 4 and in the code archive
has been corrected. Please reload them
both. (March 25, 9:30pm)
 Submission instructions for Assignment 4 are now
available on the Assignments web page.
 Assignment 4 is now posted on the Assignments
page. It is due Friday April 4 at 8pm.
 Submission instructions for Assignment 3 are now
available on the Assignments web page.
 Assignment 3 is now posted on the assignments page.
It is due on Tuesday March 18 at 3pm.
 For the midterm, you are responsible for all
material covered before the midterm in class, in
tutorial and on the assignments. However, the focus of
the midterm will be on neural nets.
 For one question on the midterm, you will need to
compute basic derivatives and use the chain rule.
 The midterm test will use the "I don't know" policy.
i.e., If you leave a question blank and simply write
"I don't know", you will receive 20% of the value of
the question.
 There will be a class after the midterm.
 The midterm test will be on March 7 in
class. It will start at 11:10am sharp and will
be 50 minutes long.
 The midterm will be closed book, but you will be
allowed one page of notes (singlesided, 8.5x11 inch)
in 12point font (or larger) and no more than 6000
characters. No other aids are allowed.
 Assignment 2 is now posted on the assignments page.
It is due on Tuesday Feb 25 at 3pm.
 Refresh your calculus with these videos
 Submission instructions for
Assignment 1 are now available on the Assignments
web page.
 Assignment
1 is now posted. It is due on Tuesday Feb
4 at 3pm.
Lectures: Fridays 11:00am1:00pm in
CC 2130
First lecture January 10; Last lecture April 4; No
lecture on February 21
Click here for a list of
all the lectures (subject to change).
Tutorials:
Wednesdays 11:00am
 12:00pm in IB 220
Fridays 10:00am 
11:am in CC 2150
First tutorials:
January 15 and 17
Click here for more
tutorial information.
Instructor:
Anthony Bonner
email: [my last name]
[at] cs [dot] toronto [dot] edu
Office: CC 3079 (UTM), BA
4268 (St George)
Phone: 9058283813 (UTM), 4169787441
(St George)
Office Hours: Fridays 2:00  3:00pm.
Teaching Assistant: Yue Li
email: yueli [at] cs [dot] toronto [dot] edu
Prerequisites:
informally: calculus, linear algebra, statistics and
computer programming
formally: CSC207H5/270H5,
290H5; MAT223H5/248Y5; STA257H5
Required Readings: There is no required
textbook for the class.
There will be one or two required papers or chapters
per week (see Lectures and
Readings). These required readings will all be
available on the web. You
may also find the following book useful, though the
mathematics can be quite advanced:
Pattern Recognition and Machine Learning
Marking Scheme:
Closed book Midterm test worth 20%
Closed book Final exam worth 40%
Four assignments worth 10% each
On all work, 20% of the mark will be for quality of
presentation, including the use of good English. The
final exam and midterm will be based in part on the
assignments and will assume that you have completed
them by yourself. Final marks may be adjusted up or
down to conform with University of Toronto grading
policies. Late assignments will not be accepted
Computing:
The assignments will all be done in Matlab, but prior
knowledge of Matlab is not required. Basic Matlab will
be taught during the first few tutorials.
Course Information
sheet: click here.
Plagiarism and
Cheating:
Honesty and fairness are
fundmental to the Univrrsity of Toronto's mission.
Plagiarism is a form of academic fraud and is
treated very seriously. The work that you submit
must be your own and cannot contain anyone else's
work or ideas without proper attribution. You are
expected to read the handout How
Not to Plagiarize and to be
familiar with the Code
of Behaviour on Academic Matters, which is
linked from the UTM calendar under the link Codes and
Policies. The following website may also be helpful:
Advice
on academic offences.
