CSCI-739-02: Introduction to Machine Learning (Fall 2018)

  RIT Department of Computer Science

CSCI-739-02 Introduction to Machine Learning (Fall 2018)

Instructor: Prof. Richard Zanibbi (Contact info)
Office Hrs: T/Th 2:00-3:20 GOL 3551
Lectures: T/Th 3:30-4:45pm GLE 3139
Teaching Assistant: Timothy Zee (Contact info)
TA Office Hrs: M/W 11:00am-12pm GOL 3650
Tutorials: (from Wk 2) Fri 1pm & 2pm Brown Hall 1110

News --- Schedule --- Syllabus --- Resources --- MyCourses


Week 15

  • EXTENSION: Project 2 is due Thursday, Dec. 13th at 11:59pm. Late submissions will be accepted until Friday, Dec. 14th at 11:59pm.
  • The final exam will be Tuesday, Dec. 18th from 1:30-4pm in GLE 3139 (our classroom). Students are permitted to bring one page of notes with things written front-and-back ('cheat sheet') to the exam.
  • Please complete the course evaluation. Comments on what worked and what could be improved in the course would be greatly appreciated.
  • Prof. Zanibbi's office hours next week (wk of Dec. 10th): Mon. Dec. 10th, 10am-12pm and Wednesday Dec. 12th, 11am-1pm.
  • Grading keys for A3 and Project 1 are now available through MyCourses. We are aiming to have both graded by Mon. Dec. 10th.
  • Tutorial this Friday will cover Project 2.
  • This week we will finish talking about LSTM/BLSTM networks, and hold a brief review for the exam on Thursday.

Week 14

  • Assignment 4 is due next Wednesday (Dec. 5th, 11:59). The assignment is on Markov Models, and does not require any programming.
  • Tutorial on Friday will focus upon Assignment 4.
  • Project 2 will be posted by Friday. Please select your group number for teams of two on MyCourses (under the "Groups" link).
  • Deadline Extension for Project 1 due to the power outages over the holiday weekend, the deadline for Project 1 has been extended to Wednesday at 11:59pm. Note: late submissions are still due by Thursday, 11:59pm.
  • We will continue talking about Recurrent Neural Networks this week.

Week 13

  • Project 1 is due on Tuesday, Nov. 27th (after the break).
  • There will be no class on Thursday due to the holiday break.

Week 12

  • Project 1 is due on Tuesday, Nov. 27th. Instructions on how to work with the nsynth dataset on 'granger' have been posted on MyCourses.
  • There will be another tutorial on Friday related to the project.
  • This week we will finish our discussion of HMMs, and start discussing Recurrent Neural Networks (RNNs). Assigned readings will be included in the lecture slides/notes.

Week 11

  • Project 1 has been released. It is due Tuesday, Nov. 27th at 11:59pm.
  • There will be a turial on Friday, going over the first project.
  • Assignment 3 extension: due Tuesday at 11:59pm. Please review the requirements of the assignment carefully before submitting your final report and code.
  • Project 1 will be released by early next week. The project will be completed in teams of two. You and your partner need to join the same group (numbered Proj1- 1 through Proj1- 11) under the "Groups" link on MyCourses. All students must complete the project in a group of two, no exceptions. Please sign up with your partner before class on Thursday.
  • This week we will be talking about probabilistic graphical model for sequence analysis (e.g., Hidden Markov Models (HMM)). Associated readings from the Barber book are provided in the lecture notes.

Week 10

  • Assignment 3 is due Sunday at 11:59pm.
  • Thursday office hrs moved to Friday, 3-4:30pm this week.
  • Friday there will be a tutorial (Questions on A3).
  • Papers on GANs have been posted on MyCourses.

Week 9

  • Assignment 3 has been posted, and is due next Sunday.
  • There will be a tutorial on Friday related to Assignment 3.
  • Keys for Assignments 1 and 2 are now available in MyCourses, including sample programs. Make sure to review both the key and the code.
  • We will continue our discussion of Conv. Networks and training networks this week; we'll try to get to Generative Adversarial Networks (GANs) next week.

Week 8

  • There will not be a tutorial this Friday. Perhaps stop by the AI@GCCIS event in the atrium, where Tim will be presenting his work.
  • We are covering Ch. 9 in the Deep Learning book this week, on Convolutional Neural Networks.
  • Assignment 3 will be delayed - we will update you when we know more.

Week 7

  • Read Deep Learning Ch. 6.1-6.4
  • This week we are starting to discuss deep feedforward neural networks.
  • There will be a tutorial on Friday.

Week 6

  • Assignment 2 has been slightly revised (to correct wording in Q2) and is due Sunday.
  • Read Section 5.11 of the Deep Learning book.
  • This week we will finish up with graphical models, and switch to talking about fundamentals for 'deep' neural networks.
  • There will be a tutorial this Friday, focusing upon Assignment 2.

Week 5

  • Assignment 2 has been posted on MyCourses. It is due next Sunday.
  • This week were are shifting to probabilistic graphical models.
  • Read Ch. 3/4 of the book on Belief Networks by Barber.
  • There will be a tutorial this Friday.

Week 4

  • Read Ch. 5 of the Deep Learning book (link to chapter).
  • Assignment 1 is due Sunday at 11:59pm. Please note that the assignment has been updated to clarify Question 1.
  • There will be a tutorial this Friday related to Assignment 1.

Week 3

  • Lecture slides and whiteboard captures are available through MyCourses.
  • There will be a tutorial on Friday (at 1, repeated at 2pm) introducing PyTorch. Tutorials are in Brown Hall, 1110.
  • This week we will be reviewing probability, distributions, and statistics.
  • Read Ch. 3 of the Deep Learning book. (link to chapter)
  • Assignment 1 will be online by Friday afternoon, and will be due next Sunday at 11:59pm.

Week 2

  • The TA will hold office hours on Wednesday this week, and give a tutorial on Friday (at two times - 1pm and 2pm). TA office hours will be in the CS Grad Lab, and tutorials will be given in Brown Hall (details above).
  • The final exam will be given Dec. 18th at 1:30-4pm in our regular classroom (GLE 3139).

Week 1

  • The course syllabus and schedule are online.
  • Assignments and projects for the course will be submitted and graded through MyCourses.
  • References and textbooks are available through MyCourses. Readings will be announced throughout the semester.