FAS - Intelligent Systems Program
The graduate program in Intelligent Systems serves as a center at the University of Pittsburgh for advanced education and research in artificial intelligence and related areas of cognitive science. Drawing on strengths from many sectors of the University, and on over thirty funded research projects, the program offers a strong, well-balanced core curriculum in the fundamentals of AI and many opportunities for advanced research and training. The scope of the program is broad, but offers concentration in specific areas such as automated diagnosis, knowledge representation, machine learning, intelligent tutoring, natural language processing and discourse, planning, case-based reasoning, and uncertain reasoning. There are especially strong connections to research groups in the Department of Computer Science, the Learning Research and Development Center, the Department of Information Science and Telecommunications, and the School of Medicine. The program also draws on associated faculty from other units, including the School of Law, the Joseph M. Katz Graduate School of Business, the Department of History and Philosophy of Science, the Department of Linguistics, the Department of Philosophy, and the Department of Psychology.
Students in the program may concentrate in such areas as:
Formal studies of reasoning and learning, including default reasoning and induction, reasoning with uncertain information, machine learning, planning, machine diagnosis, human cognition, natural language processing, and applications of these theories.
Experimental testing and validation of systems in applications such as diagnosis, inheritance reasoning, and planning.
The nature of interactions between people and information processing tools, including the interaction between a student and a machine or human tutor, between a computer user and a computer, between a database user and the database, between a professional and a diagnosis machine, and between a programmer and a programming environment.
Case-based reasoning, and applications of case-based reasoning, especially in legal domains and in tutoring applications.
Computational linguistics, and especially natural language generation and discourse.
Technologies related to the above foci, such as expert systems.
The program offers a Doctor of Philosophy degree. Guidance is also provided for students who wish to pursue joint- or dual-degrees in the program and another discipline. A Master of Science degree is available only to students who are pursuing a doctoral program in Intelligent Systems or who are enrolled simultaneously in another doctoral program elsewhere in the University. (The only exception to this rule involves the Medical Informatics track.)
Contact Information
- Main Office: 901 Cathedral of Learning
- Intelligent Systems Program
- University of Pittsburgh
- Pittsburgh, PA 15260
- Phone: (412) 624-5755
- Fax: (412) 624-6089
- E-mail: application@pogo.isp.pitt.edu
- Web site: http://www.isp.pitt.edu/
Director: Bruce Buchanan
Affiliated Faculty: ASHLEY (Law), BELNAP (Philosophy), BUCHANAN (Computer Science), COOPER (Medicine), DALEY (Computer Science), DRUZDEL (Information Sciences), FRIEDMAN (Medicine), GADD (Medicine), GRUSH (Philosophy), HIRTLE (Information Sciences), LESGOLD (Psychology), LEWIS (Information Sciences), LOWE (Medicine), MAY (Business), MOORE (Computer Science), MUNRO (Information Sciences), RAGHAVAN (Learning Research and Development Center), VanLEHN (Computer Science), WAGNER (Medicine)
Research
The faculty in the Intelligent Systems Program has a wide range of research interests, including the following: plan generation and plan recognition; resource-limited reasoning; computational models of rationality; case-based reasoning; legal information management and retrieval; philosophical logic; philosophy of social science and computer science; expert systems; hypothesis formation; medical decision-making; machine learning; application of probability theory and decision theory to medical informatics problems; bayesian belief networks; causal discovery; data mining; theory of learning algorithms; genetic algorithms; mathematical and philosophical foundations in computer science; decision-theoretic methods in intelligent systems; decision support systems for strategic business planning; development of information resources; methods for evaluation in medical informatics; design, evaluation, and implementation of computer-based tools for medical education and assessment; implementation and evaluation of clinical systems; clinical decision support; computer-based patient record systems; models of clinical communication; cognition; mental representation; neural systems; spatial cognition; classification; mathematical psychology; formal modeling and intelligent computer-based instruction of complex skills; operator modeling; visualization of constraints and information spaces; virtual environments; human-agent interaction; integrated multimedia medical record systems design and implementation; telemedicine applications of the electronic medical record; clinical, research, and educational applications of the Internet; use of goal-based task representations in multimedia workflow/project-management; supporting problem-based learning for medical teams; decision support systems in management science and in engineering; natural language processing; discourse; multimedia interfaces; neural networks; neurobiological models; data compression and error correction; intelligent tutoring systems; cognitive simulations; machine learning; constructionof decision-theoretic reminder systems; computer-assisted medical decision-making; and data accuracy in computer-based medical records. For a complete listing of specific faculty members' research interests, see the program's Web site.
Computational Resources
Faculty and students in the Intelligent Systems Program are key personnel in many well-established research projects and centers. There are well-equipped laboratories associated with research groups in Computer Science, Information Science, the Learning Research and Development Center, and the Center for Biomedical Informatics. The program maintains a laboratory for student use, which is equipped with PCs and Unix workstations. Optical-fiber based Ethernet links program machines to the campus backbone, PITTNET, which provides access to all University computing facilities, the Pittsburgh Supercomputing Center, and the Internet.
Admissions
Students interested in applying to ISP's joint-degree program (that is, those pursuing an MS degree while pursuing a doctoral degree in another program), should find details about admission to joint-degree programs on the ISP's Web page at http://www.isp.pitt.edu/. Students applying to the regular PhD program should follow the procedures detailed below.
Application Procedures
The program maintains updated information about program activities and detailed application instructions. Applicants should visit the program's Web site at http://www.isp.pitt.edu before completing an application. Briefly, an application consists of the standard Faculty of Arts and Sciences admission forms, in addition to the following materials:
A concise statement of purpose, providing information on the following points:
- 1) Objective in pursuing a PhD in Intelligent Systems
- 2) Theoretical background in relevant areas
- 3) Background in relevant tools and applications (students should list programming languages with which they are familiar, indicating level of familiarity)
- 4) Relevant practical experience, including industrial or commercial experience
Graduate Record Examinations results; applications are not complete until scores have been received.
Applicants who are not native speakers of English must take the TOEFL examinations and submit scores. Applicants from abroad should take note of the special instructions that apply. See also Graduate Admissions of International Students.
Applicants with the BA degree who are not enrolled in graduate programs at the University of Pittsburgh: These applicants should follow the normal application procedures for the PhD degree. The admissions committee is looking for students with strong research potential. Recognizing that people can come to this field in many ways, the program invites applicants from a wide variety of educational settings and disciplinary backgrounds. The committee does look for evidence of advanced standing and outstanding performance in some of the core areas relevant to the subject, including theoretical and applied computer science, cognitive psychology and other areas of cognitive science, linguistics, and symbolic programming and software engineering.
Applicants for the Medical Informatics track: Such applicants must specifically indicate their interest in this track. For additional information, please contact Dr. Gregory Cooper at gfc@cbmi.upmc.edu.
Applicants for the MS degree only who are enrolled in other graduate programs at the University of Pittsburgh: Applicants who are already in residence at the University should contact the program directly before filling out an application. Application procedures are simplified in this case. To be admitted, such applicants must demonstrate, for at least one term, satisfactory progress in their primary department. Full admission to the ISP MS program requires the completion of the equivalent of a master's degree in the parent department and approval to continue for the doctorate in that department.
Applicants who wish to apply for simultaneous admission to the ISP MS and another PhD program: Such applicants will be taken into consideration. To be admitted provisionally to the master's program, a student must be admitted to an appropriate home department by that department's admissions process. The admission will be provisional; provisional status will be removed when the applicant demonstrates satisfactory progress, for at least one term, in the primary department.
Financial Assistance
The program endeavors to provide financial support for all admitted students who require it. Support is currently available through a variety of sources, including externally supported research and training grants, University fellowships, and program funds. Further details are available on the ISP's Web page at http://www.isp.pitt.edu.
Degree Requirements
The minimal requirements established by the Graduate Faculty of the University, as described under General Academic Regulations, and any additional requirements of FAS Graduate Studies described under FAS Degree Requirements, should be read in conjunction with program-specific degree requirements described in the following sections.
Requirements for the Master's Degree
The Intelligent Systems Program offers a Doctor of Philosophy degree. A Master of Science degree is available only to students who are pursuing a doctoral program in Intelligent Systems, or simultaneously elsewhere in the University, or who already hold the MD degree or an equivalent and are specifically admitted to the Medical Informatics track.
For details on requirements for MS degree (both standard MS and the Medical Informatics track), see detail under Requirements for the PhD Degree below.
Requirements for the PhD Degree
To earn the Doctor of Philosophy degree in Intelligent Systems, a student must complete a program of study approved by an advisory committee of faculty. This program must include:
- (a) The required courses shown below
- (b) At least four advanced courses in the field of intelligent systems
- (c) An MS-level project, approved by the faculty after an oral prospectus presentation, involving significant research, design, or development work and a written report
- (d) Successful completion of a comprehensive examination composed by an advisory committee
- (e) An acceptable dissertation
Successful completion of (c) satisfies the preliminary evaluation requirement of FAS. The advisory committee may waive requirements that have been satisfied through prior university-level study. The advisory committee normally consists of three faculty. At most one of these may be from outside the program, and no more than one can be non-tenure stream. At the time a dissertation committee has been formed, the advisory committee will be replaced by a larger dissertation committee, which will supervise and approve work on the dissertation.
Course Requirements for Standard (Non-Medical Informatics) Track
This section describes the courses required for normal progress through the Intelligent Systems graduate degree programs. Check with the program to ensure that the course requirements listed below are the most current.
Prerequisite Courses (or equivalents)
- CS 1510 Design and Analysis of Algorithms
- CS 1511 Introduction to the Theory of Computation
- CS 1571 Introduction to Artificial Intelligence
- CS 1573 Artificial Intelligence Programming
- PHIL 1500 Symbolic Logic
Core Courses (required for MS and PhD)
- ISSP 2160 Foundations of Artificial Intelligence
- ISSP 2170 Machine Learning and Communication
- ISSP 3712 Knowledge Representation
Theory Courses
To complete the MS, students must take the courses listed under (A) and (B); Students pursuing the PhD must take those courses as well as the courses listed under (C):
- (A) One course in applied or mathematical statistics; acceptable courses are:
- STAT 1631 Intermediate Probability
- STAT 1632 Intermediate Mathematical Statistics
- (B) One course in theory of computation or algorithms; acceptable courses are:
-
- CS 2110 Theory of Computation
- CS 2150 Design and Analysis of Algorithms
- ISSP 3520 Theory of Learning Algorithms
- (C) One additional course; any of the courses listed above are acceptable.
Equivalents of any these courses may be accepted by petition to the MS or PhD advisory committee.
Advanced Courses
Four advanced courses, for both MS and PhD students, must be taken. For PhD students, these courses need not be taken before receiving the MS degree. In fact, one purpose of the requirement is to encourage advanced students to participate in some courses.
Master's Thesis and Doctoral Dissertation Requirements
Master's Students
ISSP 2000 Research and Thesis for the Master's Degree (used for MS project)
Doctoral Students
PhD Comprehensive Examination
ISSP 3000 Research and Dissertation for the PhD Degree
Requirements for Medical Informatics Track
This section contains the requirements for obtaining a graduate degree in the Medical Informatics track of the Intelligent Systems Program (ISP/MI).
The following curriculum assumes that a student already has training in a health-care field; if this is not so, then the faculty will select a set of courses that teach the student basic medical knowledge, and the student may take these courses as electives.
Prerequisites
It is assumed that students enter the program with the following or equivalent knowledge. If not, then these courses or their equivalent should be taken, or the student should be prepared to learn this material as self study:
Ability to program in C, C++, or Pascal
Basic computer science concepts, for example the material covered in J.G. Brookshear's Computer Science: An Overview (Benjamin/Cummings, Redwood City, CA, 1991)
CS 1541, or equivalent course about computer organization
CS 1573 Artificial Intelligence Programming
Required Course Work
The following courses are all required for the MS and PhD degrees:
- ISSP 2010 Medical Informatics Journal Club and Colloquium (offered every term)
- ISSP 2015 Introduction to Medical Informatics (offered in Fall Term)
- ISSP 2060 Evaluation Methods for Medical Informatics (offered in Spring Term)
- ISSP 2070 Probabilistic Methods for Computer-Based Decision Support (offered in Fall)
- ISSP 2240 Decision Analysis and Decision Support Systems (offered in Spring)
Two out of any three of the following courses are required for an MS degree. All three are required for a PhD degree.
- ISSP 2160 Foundations of Artificial Intelligence (suggested time: first year)
- ISSP 2170 Machine Learning and Communication (suggested time: second year)
- ISSP 3712 Knowledge Representation (suggested time: second year)
One of the following for the MS degree and PhD degree:
- CS 1501 Data Structures and Algorithms
- CS 1510 Design and Analysis of Algorithms (undergraduate-level)
- CS 2150 Design and Analysis of Algorithms (graduate-level)
- CS 3150 Advanced Topics in Design and Analysis of Algorithms
One of the following for the MS degree and both for the PhD degree:
- ISSP 2040 Introduction to Clinical Multimedia and the Internet
- ISSP 2190 Knowledge Representation and Clinical Decision Support Systems
The following is required for PhD students only:
- TA ISSP 2015 (or an equivalent course in consultation with ISSP faculty.)
Master's Students
ISSP 2000 Research and Thesis for the Master's Degree (used for MS project)
Doctoral Students
PhD Comprehensive Examination
ISSP 3000 Research and Dissertation for the PhD Degree
Course Listings
- ISSP 2000 Research and Thesis for the Master's Degree
- ISSP 2010 Medical Informatics Journal Club and Colloquium
- ISSP 2015 Introduction to Medical Informatics
- ISSP 2020 Topics in Intelligent Systems
- ISSP 2030 Selected Topics in Intelligent Systems
- ISSP 2040 Introduction to Clinical Multimedia and the Internet
- ISSP 2050 Intelligent Systems Research Seminar
- ISSP 2060 Evaluation Methods for Medical Informatics
- ISSP 2070 Probabilistic Methods for Computer-based Decision Support
- ISSP 2080 Advanced Medical Informatics Seminar
- ISSP 2140 Introduction to Parallel Distributed Processing
- ISSP 2150 Information Processing Systems
- ISSP 2160 Foundations of Artificial Intelligence
- ISSP 2170 Machine Learning and Communication
- ISSP 2190 Knowledge Representation and Clinical Decision Support Systems
- ISSP 2210 Artificial Intelligence
- ISSP 2221 Human Information Processing
- ISSP 2230 Natural Language Processing
- ISSP 2240 Decision Analysis and Decision Support Systems
- ISSP 2250 Research Design
- ISSP 2260 Visual Languages
- ISSP 2400 Intelligent Systems Laboratory
- ISSP 2510 Seminar on Artificial Intelligence and Legal Reasoning
- ISSP 2630 Foundations of Cognitive Science
- ISSP 2710 Medical Expert Systems Seminar
- ISSP 2805 Artificial Intelligence and the Logic of Discovery
- ISSP 2820 Pragmatics
- ISSP 2850 Computational Linguistics
- ISSP 2871 Art of Logic and Computation
- ISSP 2875 Logic Programming and Computational Morphology
- ISSP 2900 Graduate Internship
- ISSP 2902 Directed Study
- ISSP 2990 Independent Study
- ISSP 3000 Research and Dissertation for the PhD Degree
- ISSP 3120 Natural Language Processing
- ISSP 3180 Visual Languages and Visual Programming
- ISSP 3300 Advanced Topics in Expert Systems
- ISSP 3360 Structure and Interpretation of Computer Programs
- ISSP 3370 Artificial Intelligence in Business
- ISSP 3380 Knowledge-based Expert Systems
- ISSP 3390 Advanced Topics in Artificial Intelligence
- ISSP 3520 Theory of Learning Algorithms
- ISSP 3565 Advanced Topics in Artificial Intelligence
- ISSP 3570 Advanced Topics in Computational Rationality
- ISSP 3610 Seminar in Learning and Instructional Processes
- ISSP 3712 Knowledge Representation
|
