Minutes, IST230 Development Comm

18 Feb 2000

Frank E. Ritter ritter@ist.psu.edu

The meeting came to order at 9:04 am at UP. Lambert, Warms, Ritter attending. Others were snowed in, and were not at their desks for a phone call (Itinger checked for us).

We noted that problem-based means two things (a) to engage students in a real-world or real-world like problems, and (b) homework sets.

Items todo are in red and bold. done already are in bold only.

Agenda

Agenda as in the headings below. Please note that we are working towards a May 10/11th IST spring meeting where the committee should be able to overview the course, present a revised syllabus, note suggested module topics, and recommend a book.

Review of syllabus

We covered some topics for background and covered some topics to inform the syllabus.

CSE 260 discrete math for CS

This is background, as it is a similar but different course.

This course stresses proofs to support upper division CS algorithms courses. Rosen has been used in this course. Not an appropriate model for IST230.

Review of students

2 out of the 3 options will had a semester of calculus (business or science calculus), 1 option will have had college algebra and trig. As a first semester sophomore year course the students will have had IST110, and most of the time have had IST210 (i.e.. they will know databases and SQL), IST220 (ie networking), one CS class (ie elementary C++), so simple programming exercises are possible, 1-d array is the least upper bound of knowledge.

They won't want to take this course (initially) and may be scared of this course. To be successful, the course will need to engage students.

What course contents will be used for

Third year options. Further networks, further databases, software systems. Behavior of algorithms, variance in performance and user behavior modeling (normative and descriptive, pert, neural nets, productions systems (both cognitive and industrial))

The course will help them learn to think in a formal way.

Review of topics

The order of topics is basically good but will have to depend on textbook chosen (keep this in mind).

Other Faculty at UP who are teaching networks and databases liked this set of topics.

A few new topics/subtopics/examples

Module 1 we noted was simple, and could be taught without a book, basically as review. Teachers are likely to be able to do this. But it needs examples, e.g. use of inverse functions.

Pigeon hole principle is implied here but not in the syllabus, if fit to book it goes in module 1.

Shannon's H (to be added by a web page) not just an application of logarithms, but included as a primary definition.

Applications to be added: complex numbers/vectors for GIS.

Module 4. graphs and trees. matrix representations to move to applications (short, but not strictly speaking necessary, but probably put into to support advanced topics in Rosen like graph isomorphism). application to add: Petri nets as models of social processes and systems.

Module 5, all applications into core. (Epp has some good examples.)

What else is necessary for modeling users/systems? (Having trouble thinking of more, it appears a subset of the current syllabus.)

Review of books

One-page web pages might help to supplement any book.

The subcommittee of three looked over the books and came up with the following thoughts. We might wish to revise those thoughts, but the seemed pretty good at the time.

Worth looking at further

Washburn, Marlowe, Ryan. Discrete mathematics. Worth looking at further. See summary by Lambert to come up. [done 23 feb 00!]

Dymaccek and Sharp, Introduction to discrete mathematics. Good coverage of topics but for grammar and FSM (module 6). Worth looking at in Depth. See summary by Lambert to come up. [done 23 feb 00!]

Not worth looking at further

Grossman. discrete mathematics for computing. Not sufficient coverage, particularly sample spaces as a basis for statistics. and missing module 6. But nice-ish presentation.

Barnett, Discrete Mathematics. Nice sort of coverage, but examples and presentations are eclectic.

Kolman, Busby, Ross. discrete mathematical structures. would have to be augmented with recursion, and may be a bit intense, but not as intense as Rosen.

Rosen. Discrete Mathematics and Its Applications A little too dry, a little too complex. (Grimes via email, a strong vote for, however.)

Aho and Ullman, too much and C++ (Lambert via email)

Liu, Elements of discrete math) is out of print.

Epp - Warms didn't think this was a great book.

Grimaldi, Discrete and combinatorial mathematics. Too detailed, set of topics a bit broad.

To be looked at further

TODO: look at other books (Bottelochi has 25 books. (!))

TODO: Ritter to query Addison Wesley [done 28/2/00]

TODO: Ritter to query Prentice-Hall [done 28/2/00]

TODO: previous sponsor to present case for DIMACS modules and COMAP. Their site is http://www.comap.com.

Possible web Additions to the books

ToDo: Frank a page on Shannon's H, Caesar's cipher and Markov processes.

TODO: Tom: Epp has some examples for module 5, put these up.

With a lab?

We thought about what this would be like as a lab course with an associated lab. It looked a bit messy and we didn't see a clear line forward. Thoughts included. maybe a UNIX/Linux as a running example/exercise? And CS101 is C++ across all the campuses, so small programming projects. Oh, and the exercises could be varied by group and shared with the class as examples! Projects/homework sets outside of classes.

We discussed asking for a TA to set up a web site with examples, but the discussions were inconclusive.

Problem solving applications with respect to syllabus

Students should be put into groups with mixed abilities and mixed backgrounds and mixed IST intended options.

Crypto could be a theme: by hand, with letter counter, with auto rotator use in module 1, 3, and 5. Class exercise on cryptography: Caesar cipher. breaking it, rotating Caesar cipher [maybe too hard], PGP software use?

mean time to failure as an application/ exercise

markov diagrams, summary of human and group behavior

spanning tree to support network redundancy (could we have a network simulator to use spanning trees and to compute costs?) either hand-rolled, or off the net

Software

requirements for hardware/OS

appears varied. C++ background, but hardware is too varied.

TODO: all, submit list of equipment you could run on / with to chair for summary.

requirements for the modules

Recursion, search, exponential growth

Tarski's world. Runs on mac and windows. a blocks world that provides a way to represent logic statements, check for correctness, and tests if the statement holds for the world or not.

Next meeting, starting online

Enclosed are some action items that you should all attend to, both above and below this note. [Because we are distributed and we don't have a year, please feel free to work online, e.g., put things on the web and to use email. We will be working on a physical meeting, but may have several subcommittees of work. Similarly, if you wish to talk, please feel free to call me or each other, but if you do something substantial please share it.]

Ideas for group projects, mock up a web page?

How to get user of [user, technology and information] into this story?

TODO: All Need to push on this repeatedly.

 

A class by class topic set.

Need to get book chosen mostly.

TODO: Lambert to draft drafty list of classes by week with more detailed topics. (linked in somewhere now) [done]

Need more and more complete description of possible suggested software resources, with example URLs etc.

E.g. a network simulator to test min. spanning trees, with cost of trees and redundancy noted or computable by hand.

grammar/language demo. could use a UNIX server at UP with a guest account? need a volunteer to prepare a lab/homework to teach module 6 stuff.

 

TODO: Those not here

comment on minutes

ToDo: all

send indoor times to Rhonda