Topics in Optimization for Machine Learning Sebastian Stich

News

04.01.2022

Feedback

Dear students,

I have just uploaded feedback on report 2.

In the next days I will contact you to arrange a (zoom) meeting to provide feedback on your final presentation drafts. Ideally, these discussions should take place in 2-4 weeks from now, so that there... Read more

Dear students,

I have just uploaded feedback on report 2.

In the next days I will contact you to arrange a (zoom) meeting to provide feedback on your final presentation drafts. Ideally, these discussions should take place in 2-4 weeks from now, so that there is enough time left to incorporate the feedback.

I will invite you in groups of 2-3, so that we can have a bit of a discussion and that you also get feedback from your peers.

It would be great if you could prepare a sketch of the presentation (just bullet points - no polishing) that you can present in 5-10 minutes (max!). We will use the rest of the time for discussions, feedback and additional questions you might have on the content of the paper. Please,

  • follow the provided template structure (or some other logical structure)
  • focus on the most essential points only (maybe highlight them in color)
  • consider that the (final) presentation should be self-contained/simple enough, so that your peers can follow. Show how you will address this aspect.

-Sebastian

13.12.2021

Feedback uploaded

Dear students,

I have uploaded a brief feedback on your first report on CMS (and added the reports on the google-drive link - for internal use only - please to not share publicly).

-Sebastian

---

Evaluation matrix (based on page 5 of the presentation in... Read more

Dear students,

I have uploaded a brief feedback on your first report on CMS (and added the reports on the google-drive link - for internal use only - please to not share publicly).

-Sebastian

---

Evaluation matrix (based on page 5 of the presentation in the kick-off meeting).

1. Description of the content/main result of the paper.
2. Relation to prior work.
3. Review (comments on strengths and limitations).
4. Additional personal comments (any out of: comments on applications, future directions, judging writing quality, presentation of materials, etc.)

Each criteria is awarded 1-5 points (with the possiblity to gain max. 1 extra point in each category for outstanding performance).
The mapping of points to grades will be communicated later. In any case, a total of 20 (or above) will be mapped to the highest possible grade. 

3 - sufficient (60%)
4 - good (80%)
5 - excellent (100%)

 

29.11.2021

First Assignment due

Dear all,

- the deadline for the first report is approaching. Please check that the submission system (cms) is working so that you do not encounter technical problems shortly before the deadline.

- I also found these resources that could be helpful for the... Read more

Dear all,

- the deadline for the first report is approaching. Please check that the submission system (cms) is working so that you do not encounter technical problems shortly before the deadline.

- I also found these resources that could be helpful for the final polishing / proof reading: Section "Review content" in the NeurIPS reviewer guidelines, or the AISTATS reviewer guidelines. These are just complementary resources. If you followed the themplate from the slides this will be fine as well.

-Sebastian

01.11.2021

Topics finalized

Thanks for voting. I have now finalized the topic assignment (taking your preferences into account). You find your assigned topic in the google sheet.

Please note that the first report is due at the end of this month.

Date of final... Read more

Thanks for voting. I have now finalized the topic assignment (taking your preferences into account). You find your assigned topic in the google sheet.

Please note that the first report is due at the end of this month.

Date of final presentations: Unfortunately, the voting for the dates of the final presentation was less successful, we did not find 2 slots that work for all of you.

Somewhat popular was Tuesday & Wednesday (March 2-3, 2022) afternoon. Could the others also make this work? Please let me know of your time constraints on these days, maybe by slightly shifting the time this will work for all?

-Sebastian

 

 

25.10.2021

Course Materials

Other materials can be found on this google drive link. Including:

  • Report Template
  • Presentation from the Kick-off meeting
  • Chapter 1 of Nesterov's book (for reference)
20.10.2021

Kick-off meeting

Tentative slot for the kick-off meeting (via zoom):

October 25, 4pm-6pm

Please reach out immediately if this overlaps with your schedule(s).

Show all
 

Seminar: Topics in Optimization for Machine Learning

Optimization lies at the heart of many machine learning algorithms. This seminar teaches an overview of modern mathematical optimization methods, for applications in machine learning and data science. In particular, we will discuss the theoretical basics of stochastic optimization, scalability of algorithms to large datasets, and challenges in distributed optimization, such as for instance in decentralized or federated machine learning. We will cover a set of foundational papers, but also a selection of recent publications.

Organization

In this seminar, students will learn to present, discuss, and summarize papers in different areas of optimization for machine learning. Specifically, each student will get a single topic assigned to them, consisting of two papers (a lead and follow-up paper). Each student will

  • write a short seminar paper on the topic assigned to them, for which the two papers on the topic serve as the starting point;
  • prepare a presentation on the topic assigned to them;
  • write three short reviews on papers from a different topic, and prepare questions to ask the to the presenter of this paper/topic. The reviews will be shared among the group (in particular with the presenter of the topic).

Important Dates

  • Kick-off meeting October 25, 2021 (to be held online, via zoom).
  • The reviews (and questions) must be submitted during the semester, one review per month. You will get a short feedback after each submission. 
  • Mandatory feedback round/practice talk at least two weeks before the presentation (arrange exact time with supervisor).
  • The presentations will be organized in a block format during the semester break, in the week of February 28-March 4, 2022 (or February 14-18, 2022, dates to be fixed at the kick-off meeting). Participation is mandatory.
  • Hand-in of report: February 25, 2022, 23.59h.

Deliverables

  • 3 short reviews: (each contributes 10% of your final grade): Write a short review (max 1 page) on one of the papers (not the one that you are presenting) that addresses the following questions:
    1. What is the problem addressed by the paper?
    2. What was done before, and how does the paper improve on previous work?
    3. What are the strengths and the limitations of the techniques in the paper
    4. What part of the paper was difficult to understand?
    5. What are possible improvements or extensions of the techniques in the paper?

    In addition to your review you will have to submit 3 questions that you will ask the presenter of the paper.

  • Presentation: (40%). You will prepare and deliver a 30 min presentation (followed by 15 mins question/discussion) of the paper assigned to you. You will have the possibility to get feedback on your slides before the presentation.

  • Seminar Paper: (30%) You will write a seminar paper on the topic that you have presented. It must not be longer than 6 pages, not counting references and appendices. Note that appendices are not meant to provide information that is absolutely necessary to understand the paper, but rather to provide auxiliary material. Papers can be shorter, but in general the provided page limit is a good indicator of how long a paper should be.


Privacy Policy | Legal Notice
If you encounter technical problems, please contact the administrators