CS382:Class Notes

These are class notes for CS328: Discrete Modeling Development. They will be maintained by Kay and Charlie, but feel free to add parts we may have missed.

---

Thursday, May 7, 2009

• Rocket lab needs testing
• Predator/prey needs a third lecture
• OpenOffice 3.2 (coming out in late November) should have a bubble graph capability ("3d"), but we're still missing the ability to calculate an equation based from data points
• End Notes for the class still needs to be sorted out
• Squirrels question in Foundation unit (.1 to 82 squirrels per square acre), how do we define front campus?

Wednesday, April 8

Updating your unit pages:

• Requirements section - Please flesh these out more. Referring to something above won't help when Charlie takes these as stand alone points - it's rather to copy it from above and reword it slightly so it makes sense in this context. Don't worry about repeating yourself from above.
  • Also, "None" requirements don't need have to have something written after it.

Random:

• Change the highlight color in Firefox

Monday, March 30

• You're encouraged to use color in your units to help show importance, clarify, etc. but don't use the same colors as the reviewers! (Reviewers: red (Charlie), blue (Kay), grey/magenta (Ian))
• The template has been changed slightly in the metadata section in order to allow for lab feedback.
• For the metadata about each of the gen ed requirements, start with one of the following options:
  • Complete
  • Partial
  • None
• Follow this with a period, and space, and then your prose about how it's justified as complete or partial (none doesn't require any further prose).

Friday, March 13

• We decided to hold off on Tufte and other stuff about visual display of quantitative information about the visualization in the first couple of units, just focus on simple graphics tools.
• Everyone needs to make sure you answer your own quiz and CRC questions, please! Please put them in bold.

Wednesday, March 11

• We discussed switching static and dynamic models and decided not to.
• Everyone needs to indicate for their questions which answers are correct (please do this by making them bold)

Foundations

• Possibly too much stuff in Lecture 1 (too broad of a lecture, trying to do too much, but also too shallow)
  • Too many small things and yet also doesn't cover enough of the various "small things" that belong in Foundations, if we are going to cover a bunch of small things
• We can't say that we've taken real world to abstracting to models just based on one lecture - say "contributes to understanding of X...", not just "solid support"
• The heart and jellybean estimation isn't covered in the lab section as something they are going to do in lab, this is important to have in the lab section

Charlie's Idea:

• Some of the requirements of the gen ed requirements may be things we need to remind them over and over- along the same lines as needing to tell them multiple times that algebra is ok
  • Maybe take a giant list of stuff and say we are going to cover all of it throughout the semester, take ten minutes on average to let them see it. (Give pointer to you're going to see this here, here, and here.) Then tell them throughout the semester, remind them they saw it once before. At the end, when trying to tie it together, talk again about the big patterns about what you can do.
• List of 20 - 40 items, and each unit covers 5 - 10 of them, keep pointing them to list and telling them where we are on the list.
• This will make it easier to bring to the curriculum committee too, by saying they see things over and over from different angles and being able to say where.

Dynamic Modeling (Fire)

• Mentions agent-based modeling and systems dynamics (only covers cellular automata), should it be closer to the agent-based and systems dynamics?
  • We're thinking they shouldn't be mentioned at all, unless it can be done at a very high level.
• Look at the Interactivate: Fire model and see if there's a teaching concepts document that could help with generating questions.
• In lab, split them up to generate data (with different parameters) and then combine all the data together and see it turn into a sigmoid and talk about critical parameter, etc.
  • Rather than having all of them do all of the data generation.

Monday, March 9

• Heads up: when covering the gen ed requirements, we need to make sure to have more prose than just "covered" - we need to have prose about how each part of the requirements is covered by the unit.
• Labs will need step-by-step instructions for the students. Think along the lines of material to hand out to them with directions. Charlie is going to give more information on this soon.
  • Also need the optional and required elements for the writeup for each lab.
• Requirements for the class: assume they have some knowledge with algebra but keep in mind that we may need to remind them every once and a while how to do this kind of stuff
  • No diff equations, no calculus
  • Can do difference equations: where you are now + some change = where you are next
• Charlie is going to update the template with a place for authors, so you guys can put in your name, homepage, whatever

To Do for Charlie:

• We might use the gym for the rocket launching lab, so Charlie needs to check if it's possible to schedule the gym during the day.

To Do for Class:

• For Wednesday, everyone should read through all of the units with some critical eye (besides your own). Don't embed comments but take notes and come to class prepared to talk about them.
• There will be an article to read over break.
• Friday after break, the labs will be do, as well as a layout cleanup.
• Second Friday after break, everyone is going to go through and do somebody else's lab. This means preparation and setup for the lab has to be done by then!

Monday, March 2

• "Lead" on a unit will be given up to 30 points for the unit, and second person only 15 points
  • To help motivate folks to sort out who does how much work for each unit
• Modeling disease became predator-prey (after some swapping of people), and Charlie is going to get back to them about an existing model already built with curriculum
• Nate will get one and a half weeks
  • Matt's predator prey will be a week and a half and then show them (run for them or have them run) the predator prey model agent-based, and then work on the systems dynamics
  • Nate gets two labs, Matt gets one

Wednesday, February 25

• When we have them do a wiki writeup, give them a PDF of what it should look like as well as a guide to wiki syntax, but have them discover how to use it and replicate the image by themselves.

Friday, January 16

Initial thoughts from the homework reading:

• crowd sourcing - potentially we could do class sourcing with discrete modeling or something else
  • maybe something in Second Life (SL) as a virtual crowd?
  • maybe using an existing large scale model in SL?
• scavenger hunt
• make impressions on people
• high level of engagement in activities
• important points from McGonigal about games we should incorporate:
  • satisfying
  • part of something bigger - make people want to keep going, ie also tie into (mention) other large scale projects - also see next point
  • experience at being good at something
    • want students to try and succeed on their own, but get enough feedback to know if they are heading in right direction

High level goals for the course:

• use computerse to model the world and show that they're useful for the rest of students' careers
• CS is cool/useful

General points to consider for the class:

• We could look at what other liberal arts sciences classes are covering
  • Make sure we're geared also towards potential CS recruits (some non-natural sciences majors) as well as non-science majors
  • Open people's eyes to see what CS can do
• Do we have time to explore some of the technical CS behind the projects?
• What general CS principles do we want to convey?
  • Basic foundations (like abstraction, algorithmic thinking, etc.)
• Should we base units on the science or the tools/methods?
• Possibly multiple parts to assignments, or different options to chose from. Want to challenge everyone at different levels (but don't want them some noticeably easier than others, just different interests)
• Charlie needs to be able to give lots of feedback in an easy way for him to do
  • Possibly many TAs, possibly some specific to a given unit taken from a different department (ie biology student/professor for biology unit, etc.)

Wednesday, January 14

About the class:

• We will be designing a new class "in silico"
• new class will be offered the first time next spring
• geared towards first year students
• lots of this already developed, we'll be selecting the best parts of which ones for this specific course at this specific college

Themes we're designing for:

• quantitative reasoning
• model development and use
• validation and verification
  • Did I solve the right problem? Did I solve the problem correctly?
• estimation
• visualization
  • data -> information -> knowledge
  • harder to do as go further to the right
  • visualization is one way to make it easier to get more from just data
• mostly the natural sciences, possibly some art
• using tools (spreadsheets, models, make your own or pre-made)

Methods we will use:

• inquiry based learning find out how to solve a problem and document it, and describe what learned from it
• scaffolded - provide an empty framework for students to work through it in multiple ways
• metric system exclusively
• auto-magic grading?
  • Good feedback is important, possibly part of this could come from a machine

Units/modules (each probably week to two weeks) might include:

• reading
• lectures/discussion notes
• lab

Potential Units/Modules:

• "Seeing Around Corners" - an article about race behavior using Agent-based models
  • Lunch rooms, neighborhoods, etc.
• Possibly a unit with "sensor nets"
• Possible energy unit - EEAP, wind, solar
• Measuring - area, volume, count
• Ground water - wet lab, analytical, in silico
• Genomics
• Measure gravity (as between the roof of Dennis Hall and the ground)
• Something requiring lots of computational horsepower (maybe?)
  • Maybe just mentioning and not an entire unit
  • Maybe tied into the genomics unit or chemistry
• Chemistry, possibly forensic

Other Thoughts:

• We may find things that are really cool but that don't fit into this particular class we're designing. However, they may be inserted into various places in the CS curriculum - in POCO, ACS. We should capture them somewhere and Charlie will come back to them another time.