Predictive Modeling Isn’t Magic – Technology Review.

Here’s an interview with someone who does predictive modeling for a living. During the interview he discusses **agent modeling**, which models systems by creating a story in which agents interact in relatively simple ways with their neighbors and instructing a computer to calculate the net effect of many such interactions. The world can be *modeled*–represented to human beings as rich analogies–in this way. Every middle and high school student can grasp this.

At the recent Supercomputing Education Program in New Orleans, Bob Panoff of the Shodor Educational Foundation advanced his campaign to open up the world of computational modeling first for K-12 teachers, and then for students. Here’s a screengrab from his outline for one of those talks, all of which are fully available online:

Agent modeling can be introduced in K-12 in fairly simple ways. At the SC10 Ed Program, Agent Sheets was introduced to K-12 educators as a way of inviting students into computation through agent modeling. Here’s a link to my notes from that event, walking through the basics of modeling in Agent Sheets using the frog-and-the-prince story. In that story, a (red) witch that turns a (blue) prince into a (green) frog, and a (pink) kissing princess that turns the frog back into a prince. Change the witch to a bacteria, the princess to an antibiotic, and the prince into a patient, and you’ve begun to model certain kinds of disease transmission. Add many of each kind of agent (witch:princess:prince::bacteria:antibiotic:patient), and you’re modeling (at first, quite simply) the spread of an epidemic.

These notes are embedded in a computationalscience.michianastem.org wiki, which currently includes a copy of Shodor’s useful list of resources and my notes on Agent Modeling. As you might discern from my choice of venue for those notes, it is my hope that attention to computational science becomes a regular part of Michiana STEM culture, even K-12.