Artificial Intelligence Research and Education

PhD Qualifying Exam

The AI Depth Exam is designed to test for both a general "breadth" knowledge of AI, plus a deeper specialized knowledge of one particular sub-area within AI. The current sub-areas tested for, and the corresponding courses, are: machine learning (CS 760) and computer vision (CS 766).

Students will be required to specify their sub-area when they sign up for the exam, and then answer corresponding questions in that sub-area, as described in detail below. Passing the exam will then be considered as demonstration of competence in that particular sub-area, so that except in unusual circumstances, the student will be expected to pursue Ph.D. research in the area chosen.

Students will be prepared for the AI Exam if they have mastered (a) the material presented in an introduction to artificial intelligence course such as CS 540; (b) the material presented in at least two of 731, 760, 766, 776, and 769; (c) the Basic and the Advanced Readings associated with one of CS 760 and 766; and (d) the Basic Readings for at least one additional class chosen from CS 731, 760, 766, 776, and 769. Students should note that the Advanced Readings go beyond the material covered in the associated course. The introductory AI material (CS 540) will not be tested in separate questions, but knowledge of this subject matter is likely to prove necessary when answering the graduate (700) level questions.

The AI Examination will consist of questions in six separate sections. There will be two "basic" questions in each of sections B731, B760, B766, B776, B769, and two "advanced" questions in each of sections A760 and A766. To pass the examination, students will have to answer satisfactorily both of the questions in one of the A sections (A760 or A766); both of the questions in the B section (B760 or B766) for the same sub-area; and two additional questions from any of the other B sections. The two additional questions can be but do not have to be from the same section. The A section chosen must correspond to the sub-area for which the student is signed up.

The advanced questions focus on principles and concepts, and the student is expected to be able to solve new and original problems which require adapting the material covered in the reading lists below. The questions answered by the student will not in general be weighted equally. In particular, students will not pass unless they have demonstrated competence in their chosen specialty.

Reading Lists

Previous AI Qualifying Exams

Postscript or PDF versions of old AI qualifying exams are on-line at the department's archive of Ph.D. Qualifying Exams.

This page was created by palmeran@biostat.wisc.edu