Requirements of the PhD Program

Breadth Requirement

Breadth courses are divided into three groups as follows:

  • AI and HCI
    • 412 Human-Computer Interaction
    • 444 Logical Foundations of Artificial Intelligence
    • 446 Mathematical Foundations of Artificial Intelligence
    • 447 Natural Language Processing
    • 448 Statistical Speech and Language Processing
    • 449 Machine Vision
  • Systems
    • 453 Dynamic Languages and Software Design
    • 454 Programming Language Design and Implementation
    • 455 Software Analysis and Improvement
    • 456 Operating Systems
    • 457 Computer Networks
    • 458 Parallel and Distributed Systems
  • Theory
    • 484 Advanced Algorithms
    • 486 Computational Complexity

The breadth requirement is satisfied by two courses from each group as follows:

  • 446 Mathematical Foundations of Artificial Intelligence taken during the first year of graduate study, and one other AI and HCI group course  taken during the first or second year;
  • 456 Operating Systems taken during the first year of graduate study, and one other System group course taken during the first or second year;
  • 486 Computational Complexity taken during the first year of graduate study, and 484 Advanced Algorithms taken during the first or second year.

In addition to breadth and other coursework, first-semester PhD students will take 400 Problem Seminar.  The first year of graduate study will therefore be as follows:

Fall: 400, 456, 486, one other course;
Spring: 446, two other courses.

PhD students are expected to earn a B or better in all breadth courses.  Students who have already taken courses that are equivalent to one of our required breadth courses may petition to substitute a more advanced course in the same general area of computer science.

Students often wonder if they should try to complete the breadth requirement during their first year in the PhD program, or if they should take fewer courses in order to allow time for research.  The program is deliberately designed to allow students the flexibility to pursue either strategy.  If you arrive at Rochester knowing who your research advisor will be, and your advisor has an appropriate research project for you to begin to work on, then it is fine to take just three or four formal courses your first year.  On the other hand, if you are not ready to jump into research mode, then it is a good idea to take all seven breadth courses the first year.  Finally, if you joined the PhD program after majoring in a subject other than computer science, or for any reason have significant "holes" in your undergraduate education, you should work with your advisor to design a sequence of courses for your first two years that include some additional undergraduate courses that will prepare you for the breadth courses; such a modified course of study must then be presented as a petition to the faculty for approval.

Qualifying Requirements

In addition to the breadth requirements, PhD students will qualify for a research area.  All research areas require writing and defending an area paper during the second year of graduate study.  A specific due date in April for all area papers will be announced by the graduate coordinator each year.  A master’s degree is offered to those PhD-program students passing the area paper with an appropriate level of performance or, in special cases, passing an alternative comprehensive examination by given two or more faculty members appointed by the Chair.   Area specific requirements are as follows:

  • AI: The student should form a committee of three or more AI faculty member by the time the area paper is due, and schedule a time and date for the exam so that all committee members can attend.  All other AI faculty members are also encouraged to attend the exam.  The paper should include a survey of a research topic and initial original research.
    • The defense begins with the student providing a 20 minute overview of the area paper, followed by an hour of intensive questioning by faculty members.  Questions from faculty members will include general questions about AI, and in particular any material from the AI breadth courses taken by the student, in addition to questions about the area paper.  The student is responsible for preparing him or herself for such general knowledge questions.
    • By the end of the 4th year, the student must complete at least two additional courses from among the following:
      • CSC courses numbered 400 or higher that are in AI or HCI
      • BCS 505: Perception and Motor Systems
      • BCS 512: Computational Methods in Cognitive Science
  • HCI: The area paper may be a survey of a research topic, original research, or a combination of both.
    • Following a public presentation of the paper, the student will answer questions from the committee about the paper and about any topic in HCI that is broadly relevant to the research topic.
    • By the end of the 4th year, the student must complete 3 additional courses as follows:
      • At least one in experimental design and analysis: STT 422: Design of Experiments, BCS 510: General Linear Approaches to Data Analysis I, CSP 519: Data Analysis: General Linear Applications;
      • Two additional courses from the among the following: (i) CSC courses numbered above 400; (ii) BCS 505: Perception and Motor Systems, BCS 511: Behavioral Methods; (iii) the courses on experimental design and analysis listed in the previous bullet point.
  • Systems: The area paper should be original M.S. level research and a survey of related research topics.
    • Following a public presentation of the paper, the student will answer questions from the committee about the area paper and the papers listed in the systems area reading list, which includes subjects in computer architecture, programming languages and systems, operating systems, and parallel and distributed systems.
    • By the end of the 4th year, the student must complete at least 2 more courses numbered 400 or higher that are in Systems areas.
  • Theory: The area paper and defense are organized as follows:
    • The student submits an area paper reflecting research ability;
    • Approximately 2 weeks before oral exam, the student is given aset of “take home” papers;
    • Approximately 3 hours before the oral exam, the student is given a written set of “morning questions”;
    • The area defense consists of a 2 hour oral exam, covering the area paper, take home paper, morning questions, and other area-related questions.
    • By the end of the 4th year, the student must complete one additional course, 487 Advanced Modes of Computation. A more detailed written description of the logistics of this process is available from the Graduate Coordinator or any theory faculty member.


Plenetary Advising Session (PAS)

Twice each year the faculty as a whole meets and discusses the progress of every student in the PhD program.  The following week, each student receives a letter summarizing the faculty's view of their progress, congratulations for accomplishments, and advice for improvement.  Students who are in danger of not being in good academic standing, for example, by poor performance in course work (grades lower than a B) or failure to show effort in research, may be required to take remedial action, such as taking an additional course or meeting a schedule of research goals.  Students who do not respond to PAS mandates may be dismissed from the PhD program.