Maria K. Cameron

AMSC/CMSC 663, Fall 2020

Classes: Tue-Th 12:30--1:45 via zoom

Instructors:

• Maria Cameron, email: marikc@umd.edu
• Radu Balan, email: rvbalan@cscamm.umd.edu

Course website:

• umd.instructure.com mirrored at
• https://www.math.umd.edu/~mariakc/advanced-scientific-computi.html

Prerequisites

• AMSC/CMSC 660 and 661 or similar scientific computing/numeral analysis graduate classes

Course requirements

• Each student must have a faculty advisor (who is not an instructor of this course)
• Deliverables:
  • Written documents and oral presentations (see Section Presentations at proposal documents below)
  • Project proposal (5-7 page document, due Oct 8.)
  • Proposal presentation (20 min oral presentation)
  • Mid-year report (an edition of project proposal plus current results, due Dec. 10)
  • Mid-year presentation (20 min oral presentation)
  • Code posted on GitHub (if appropriate) and accompanied with documentation (see Section Code requirements) (due Dec. 16)

Grading

• Projects will be graded on the quality of
  • code and documentation for it;
  • written documents: project proposal and mid-year report;
  • oral presentations: proposal presentation and mid-year presentation.
• Oral presentations and written documents should reflect:
  • critical thinking,
  • ability to formulate and achieve research goals,
  • ability to identify and overcome difficulties,
  • newly acquired technical skills.
  • Note that a solid effort without significant technical advances will not result in a good grade.

Code requirements

• Code should be well-organized, clean, loosely coupled, and extensible
• Code should be tested and validated.
• Code should be documented, well-commented, and accompanied with a user-friendly guide.
• Code should be distributable by GitHub if appropriate.

Oral and written presentation requirements

• Project proposal and presentation should include the follwing components:
  • Background on the problem being addressed
  • Why is this problem important?
  • What are the state-of-art methods for solving it? Provide refs.
  • Project goals: what are you hoping to achieve?
  • Approach. How will you achieve these goals? What components will need to be implemented to get there?
  • Describe specific algorithms and how they will be implemented.
  • Describe hardware/software platform you target. What programming languages will be used?
  • Validation methods: how you plan to test your code.
  • Deliverables: specific components of the code you plan to develop.
  • Milestones and a rough timeline.