Artificial Neural Networks and Computational Brain Theory Group

Computer Science Classes

• CS343 - Introduction to Robotics
  AI Faculty
  Same as ECE370,GE370.
  Opinions on this class? looks like almost entirely engineering, with noactual AI component. Still if you intend to work on robotic control, probably a good foundation.
  
  
• CS346 - Pattern Recognition and Machine Learning
  Prof. Dan Roth (as of last offering)
  In the past, this class was "was an excellent opportunity for grads &undergrads to try their first presentation; final project, no exams; homeworkconsisted of writing many paper reviews; a good opportunity to read some goodpapers." However, begining Spring 98, the class will be taught by Dan Roth,and is to be a formal class w/o student presentation. Should be no less valuableas a rigorous overview of methods in Machine Learning.
  
  
• CS347 - Knowledge-based Programming
  AI Faculty
  Opinions on this class?
  
  
• CS348 - Introduction to Artificial Intelligence
  AI Faculty
  Same as ECE348.
  "a very introductory AI class; might want to try to skip it if you don't need it."
  
  
• CS442 - Artificial Neural Networks
  Prof. Sylvian Ray
  "the neural network class at the university - if you are interested inartificial neural nets and are a student here, you should definitely take this."last we heard: no exams; some homeworks; final paper/project.
  
  
• CS443 - Computer Vision
  Same as ECE 449.
  Prof. Narendra Ahuja
  "a good class on current approaches to machine vision; graduate students wererequired to present a paper of their choosing (during second half of thesemester)." last we heard: no exams; some (tedious) homeworks & programming projects.
  
  
• CS444 - Design of Computer Problem Solvers
  Same as ECE 444.
  Opinions on this class?
  
  
• CS448 - Computer Models of Cognitive Processes
  Same as ECE 448.
  Prof. Gerry DeJong
  "Title is somewhat misleading, in that little time is spent on cognitiveissues, focus is primarily on symbolic AI. However, a good opportunity to readsome good papers, and a great opportunity to engage in great discussions. Alsoyou'll receive an important introduction to research methodology in AI, from afantastic professor (DeJong)."Last we heard most grads required to present a paper, some did final paperinstead; homework consisted of writing many paper reviews and programmingprojects (looks like Java rather than LISP begining fall 1998).
  
  
• CS491 sec. SRR - Artificial Neural Networks and Computational Brain Theory
  Same as NEURO491 (avail. soon)
  Profs. Ray and Anastasio
  1/4 to 1/2 Unit.1/2 Unit credit will entail presentation of a research report.
  Pre-requisites: Graduate standing or consent of instructor
  ed. note: this class was created as an adjunct to the ANNCBT meetings, andso naturally comes highly recommended by us. For more information see ourpage on the adjunct seminar.
  

Neuroscience Classes

• NEURO303 - Introduction to Neurobiology
  Same as BIOL303.
  "a thorough, biologically detailed tour of the nervous system, stressing basic biological processes, rather than higher level functions; a large lecture class, mostly undergrads; several multiple-choice exams; a good introductory class on the biological foundations of the nervous system."
  
  
• NEURO304 - Cognitive Neuroscience
  Same as PSYCH304.
  Prof. Marie Banich
  "a great grad/undergrad class, with lively discussions of provocative cognitive and computational issues in neuroscience; taught by the best neuropsych professor in the dept.; three take-home exams"
  
  
• NEURO317 - Methods in Computational Neuroscience
  Same as BIOEN317, BIOPH317,PHYSL317.
  Profs. Anastasio and Nelson
  "excellent class (almost all grads), some programming in matlab; homework and labs; no exams. one midterm project; Class is half single-neuron modeling, and half small-network study; first half of semester gives you a thorough lesson on how abstract ANN models are, and second half is a great opportunity to read current papers and see how to interpret larger scale systems."
  There is also a homepage for this class. Note that this class is taught on alternateyears; worse yet, Anastasio is on sabatical for 98, and Nelson has plans tosabatical in 99, so it remains to be seen whether this class will be taughtanytime soon.
  
  
• NEURO410 - Advances in Psychobiology: Introduction for Graduate Students
  Same as PSYCH410.
  Faculty taught
  "cannot be recommended by me for official registration; two long take-homeexams (20 pages each) on topics as varied as the retina and the effect ofstress, with no flexibility regarding exam topics - a scheme which is perfectif you enjoy spending huge chunks of time on matter which has no relevance toyour research. might be of some value to the student with not a lot ofbackground in neuroscience, and is looking for a research area, and offers theopportunity to hear from a variety of neuropsych faculty; two 2hour classes byeach prof. mostly a quick tour of the basics - very little by way of theprofs. current research. Could be useful as a class to sit in on withoutregistering for credit."
  
  
• NEURO411 - Advanced Physiological Psychology
  Same as PSYCH411.
  Faculty Taught
  

  Official Blurb:
  Systematic study of the psychology of attention, including focused and dividedattention, dual-task performance, attention and memory, attention andautomatization, and skilled performance. The emphasis is primarily theoretical,focusing on current approaches and the historical developments that led tothem.

  Prerequisites: Graduate standing in Psychology or consent of instructor. 1/2 or 1 unit.
  Credit: 1/2 or 1 unit.
  
  
  
• NEURO416 - Neurophysiology Laboratory
  Same as PHYSL316.
  Prof. Nelson
  Same as Physiology 416
  This class may be a little low-level and detailed for most ANNCBT people.As with most classes that Mark Nelson teaches, there is a very nice web site for this class,detailing the syllabus and what is expected from students.
  
  
• NEURO487 - Human Neuroscience
  Same as CSB487.
  Prof. Weyhenmeyer
  Same as Cell and Structural Biology 487.
  does any have information on this (new?) course? Its home department is CSBwhich suggests that the focus will be on fairly low-level (cellular,physiological) descriptions.
  
  

Psychology Classes

• PSYCH404 - Theories of Attention
  Professor: Logan
  

  Official Blurb:
  Systematic study of the psychology of attention, including focused and dividedattention, dual-task performance, attention and memory, attention andautomatization, and skilled performance. The emphasis is primarily theoretical,focusing on current approaches and the historical developments that led tothem.

  Prerequisites: Graduate standing in Psychology or consent of instructor. 1/2 or 1 unit.
  Credit: 1/2 or 1 unit.
  opinions on this class?
  
  
  
• PSYCH421 - Knowledge Representation
  Professor: Brewer
  

  Official Blurb:
  Same as Neuroscience 411. Detailed examination of the physiological mechanismsin behavior; emphasis on research methodology and contemporary literature inthe physiology of motivation, learning, perception, and emotion; and includeslaboratory demonstrations and problems.

  Prerequisites:Twelve hours of psychology, including Psychology 311 or equivalent.
  Credit: 1/2 or 1 unit
  opinions on this class?
  
  
  
• PSYCH471 - Proseminar in Cognitive Science
  Same as ANTH470, CS449, LING470, PHIL470, EDPSY471.
  Prof. Gary Dell
  Same as Computer Science 449, Educational Psychology 471, Linguistics 470, andPsychology 471, Anthropolofy 470.
  "A good graduate course on a wide range of topics - good opportunity fordiscussion. Very interdisciplinary. Final assignment is to write a long paperon subject of your choosing. Should be considered mandatory for ANNCBT types.
  
  

Other Classes

• CE 498 NN - Civil Engineering Special Topics: Neural Networks in Engineering Applications
  I wonder sometimes if there is a concerted effort to hide interesting classes. If anyone has taken this class, i'd like to hear about it.
  
  
• ECE/BIOENG 375 - Modeling of BioSystems
  Looks interesting and very eclectic. Any one take this?
  
  
  
• ECE437 - Fundamentals of Speech Processing and Recognition
  Prof. Levinson
  

  Official Course Desciption: Introduction to the theory and techniques in speech processing and recognition; includes speech production model, spectral analysis, pattern comparison techniques, hidden Markov models (HMM), and HMM-based automatic speech recognition; also includes computer laboratory.
  

Seasonal Seminars

• CS491Sec.DNR - Seminar on Learning and Knowledge
  Prof. Dan Roth
  

  Instructor's Description:
  The seminar will meet to discuss current topics in learning and, in particular, topics that have to do with learning and knowledge. The idea is to present recent (and some not so recent) papers on the acquisition of large knowledge bases, learning methods that incorporate knowledge, relevant learning models, and other topics in intelligent inference that are related to learning.
  A preliminary list of papers include topics such as learning relational representations, relational Bayesian networks, Cyc, learning for abduction etc. I am open to suggestion of papers on these topics that people would like to discuss.
  
  Prerequisites: Students from all areas of CS or the Cognitive Sciences are welcome. Some background in AI as well as in Theory or Computational Learning (or the willingness to put time into it) will be needed.
  OFFERED: Fall 1998
  First Meeting: Thursday, Aug. 27, 2pm, DCL 2222. Please contact me if you cannot come to the first meeting. The main goal will be to set up time that is convenient to all the participants.
  

• CS497Sec.DNR - Machine Learning and Natural Language
  Prof. Dan Roth
  

  OFFERED: Fall 1998
  
  

• CS497Sec.DJK - Pattern and Object Recognition
  Prof. Kriegman
  

  OFFERED: Fall 1998
  
  

• GE485 - Genetic Algorithms in Search, Optimization, and Machine Learning
  Prof. Goldberg
  

  OFFERED: Spring 2000
  
  (based on the Addison-Wesley, 1989, book of the same name) on line in (1) Web format, (2) computer CD-ROM format, and (3) video tape format. The course web site, containing a course syllabus, introductory information, and a free 50-minute preview lecture are available on line at http://www.engr.uiuc.edu/OCEE/webcourses/ge485/ The free lecture entitled Darwin-in-a-Box: How Computers are Catching Criminals, Designing Jets, and Programming Computers previews the instructor's lecture style and the course's delivery technology. The course may be taken for credit (or not) through the University of Illinois at Urbana-Champaign or the National Technological University (NTU). NTU students should contact their NTU site coordinator for more information. Other interested parties should contact Linda Krute at l-krute@uiuc.edu or 1-800-252-1360, extension 36634 for further information.

• ECE497Sec.LEV - Computational Models of Spoken Language
  Prof. Levinson
  

  Instructor's Description:
  This course covers mathematical models of linguistic structureand their implementation in computational algorithms used in automaticspeech understanding and speech synthesis. Statistical and automatatheoretic techniques are studied allowing a quantitative description ofacoustic-phonetics, phonology, phonotactics, lexicons, syntax and semantics.Students will use the methods to build components of a speech understandingsystem.
  1) Introduction 1 hours
  History of Speech Recognition and Synthesis; Review of basic principles of digital speech processing; Description of unsolved problems
  2) Review of the linguistic hierarchy 3 hours
  Acoustic-phonetics, phonology and phonotactics; Morphology and lexicon; Prosody, Syntax and Semantics;3) Review of Syntactic pattern recognition 4 hours
  Sequence alignment and matching by dynamic programming; Hidden Markov models; Formal and stochastic grammars; Formal and stochastic automata
  4) Grammatical inference by parameter estimation 4 hours
  EM algorithm; Baum algorithm; Event counts and dynamic programming algorithms; Minimum Error Probability and other objective functions
  5) Probabilistic parsing algorithms 6 hours
  Instantaneous maximum likelihood state methods; Maximum likelihood state sequences; CKY algorithms
  6) Representations of segmental structure 2 hours
  Context sensitive representations of phonology; Markovian approximations to phonotactics
  7) Grammar 6 hours
  Finite state subsets of Natural Language; Markovian approximations to natural language; Context free and LR(k) formulations; Corpus based methods and SVO models; Information theoretic measures of grammatical constraint
  8) Semantics 6 hours
  Turing machine models; Formal logic methods; Graph theoretic methods; Linking semantics to grammar
  9) Speech understanding and speech synthesis systems 3 hours
  Architectures; Implementations; Performance
  10) Research topics 3 hours
  Incorporation of prosody; Pragmatics and Discourse models; Conversational systems; Synthetic vs. adaptive methods; Grounded semantics vs. abstract symbolic processors
  11) Discussion of student projects 7 hours
  Texts: Rabiner and Juang, Fundamentals of Speech Recognition (review only); Sproat, Multi-lingual Text-to-Speech Synthesis; Alshawi, The Core Language Engine; selected journal articles
  Prerequisites: ECE437, ECE434
  Recommended: ECE394
  Credit: 1 unit
  Lecture hours: 3 hours/week
  OFFERED: Fall 1998
  

---

You can also look for information in the official uiuc course catalog, timetables, or incomplete list of classes with web resources.

---

Have an opinion about one of the classes listed above or want to suggest a new class?
Let us know about it. (opinions will be posted anonymously).

---

Last updated Aug 26, 1998.