• People
  • Courses

Deep Learning, Spring 2015

DS-GA-1008

Syllabus (aka required reading)

Lecture: Mondays, 7:45-9:35pm, in MEYER 102
Tutorial: Wednesdays, 7:10-8:00pm in CIWW 109

Instructor: Yann LeCun - yann [ at ] cs.nyu.edu
Teaching Assistant: Christian Puhrsch - cpuhrsch [ at ] nyu.edu
Please only send emails about personal issues.
Please prefix all your emails with the following tag:
[DS-GA-1008 YOUR_TEAM_NAME]
with the team name as found below on the list. Otherwise the email won't reach us at all.

Prerequisites: General machine learning course such as DS-GA-1003. Highly recommended: General understanding of Unix and proficiency in at least one programming language.

Piazza: Piazza
You probably also want to read through this

Teams: List of team names, Kaggle names and performance (ranking) across assignments. Teams need to communicate their setup within the first week with an email of the following content.

Title:
[Deep Learning YOUR_TEAM_NAME] Introducing YOUR_TEAM_NAME
Body:
YOUR_TEAM_NAME
YOUR_TEAM_LEADER - name, nyu email address

YOUR_TEAM_MEMBER_1 - name, nyu email address
YOUR_TEAM_MEMBER_2 - name, nyu email address

Team leaders are responsible for managing the team’s submissions on Kaggle and submitting the team’s assignment.

Schedule

Video lectures

Evaluation:
10% - A1
20% - A2
15% - A3
15% - A4
40% - Final Exam - May 18

Assignments:
Assignments are due at the beginning of the lecture. An assignment is graded based on Kaggle performance (competition), write-up and code.

Version control: Github

HPC Guide - please do make use of this cluster, if you want. GPUs will significantly speed up your code. Don't forget to read this guide on how to use CUDA with torch.

Torch + Utilities

Lateness policy:
Homeworks are due at 7:45pm (Lecture) or 7:10pm (Tutorial) sharp. Lateness penalties are: 1-10 mins = 5% of total points; 11-30 mins = 7% of total points; 31 mins - 24 hours = 10% of total points; longer = (10% of total points) x (number of days late, rounded UP). The Kaggle competition closes at the same time and there are no late submissions. Notice, that times on Kaggle are in UTC. We are in EST.

Remarks:
- Every assignment should be done in a group of two. However, you are allowed to form groups of size up to three.

- A group name must be chosen and a group leader must communicate her/his Kaggle name to the TA within the first week (registered on an nyu email address). The link to the Kaggle competition will be posted with the assignment. The required detail per write-up may vary across assignment and will be specified as well. Movements between groups must be communicated to the TA via email and can only happen within the week an assignment is due.

- Exams will be closed book, closed notes.

- The teaching staff will use piazza to send out announcements. Please check it regularly.

- In order to pass the course, you need to get at least 30% of the possible points on the final exam. The final exam is designed to test your knowledge of the assignments and then ask other theoretical questions.

On academic integrity:
We expect you to try solving each assignment within your own team. However, when being stuck on a problem, I encourage you to collaborate with other students in the class, subject to the following rules:

- You may discuss a problem with any student in this class, and work together on solving it. This can involve brainstorming and verbally discussing the problem, going together through possible solutions, but should not involve one student telling another a complete solution.

- Once you solve the homework, you must write up your solutions within your own team, without looking at other people's write-ups or giving your write-up to others.

- In your solution for each problem, you must write down the team names of any team with whom you discussed it. This will not affect your grade.

Please also follow the GSAS regulations on academic integrity found here http://gsas.nyu.edu/page/academic.integrity

Last year's website: http://cilvr.nyu.edu/doku.php?id=courses:deeplearning2014:start - Good source of links to papers.

All product names used in this website (www.kaggle.com, www.piazza.com) are trademarks of their respective owners, which are in no way associated or affiliated with NYU.

 
 
/srv/www/cilvr/htdocs/data/pages/courses/deeplearning2015/start.txt · Last modified: 2015/05/30 04:44 by cp
Recent changes RSS feed Creative Commons License Valid XHTML 1.0 Valid CSS Driven by DokuWiki
Drupal Garland Theme for Dokuwiki