CAAM 654: Sparse Optimization

Monday and Wednesday, 4:15PM-5:30PM, Duncan 1042.
Instructor: Wotao Yin / Co-Instructor: Ming Yan

Final Project

  • About CAAM 654

    • Enrollment 21: 18 for credit, 3 audits; 1 from U.Houston

    • Majors: CAAM, ECE, STAT, Applied PHYS

  • Goals

    • Learn something interesting or even produce somethong cool.

    • Find a good partner and enjoy collaboration.

    • Sell what you learn or create; impress the class!

  • Projects

    • Either pick some papers from the course website

    • Or, propose your own topics

  • Places to find interesting topics

    • Repositories such as optimization-online.org, dsp.rice.edu/cs, CAAM tech reports, UCLA CAM tech reports

    • List of Top-Downloaded papers

    • Personal websites and blogs

    • Recent workshops and conferences (NIPS, ISMP, SIAM Imaging, Allerton, ICASSP, and so on)

    • Talk to people

  • What next?

    • Pick up the background

    • Read papers and lecture notes, watch talk videos (if they are available)

    • Try it out (download the code or write it yourself), reproduce simulation results

    • Prepare your presentation (don't start last day!)

    • Good to be ambitious, planned, disciplined, and collaborative

  • Milestones

    • End of September: team and topic, plan and goals, Owlspace submission required

    • Mid October: interim report, 1 page or a few slides, Owlspace submission required

    • November: presentation starts, 2-3 teams per meeting

    • End of semester: ZIP of your report or slides, Owlspace submission required

  • Presentation requirements

    • Be correct

    • Give clear background and motivations

    • Summarize the method and its novelty

    • Show applications and simulation results

    • Discuss open questions and future work

    • Properly credit people

  • End of semester submission. The ZIP file should include

    • Names of team members

    • Papers, slides, and/or codes downloaded

    • Your presentation slides

    • Your codes if the paper is about an algorithm

    • (optional) a report

  • About collaboration

    • You can choose to work alone

    • Find a responsible collaborator with complementary skills

    • Establish consensus on the plan and goals early

    • Set up internel deadlines