The MASSIC Project (“Multi-Agent Systems that Simulate Intelligence and Cognition”) is a series of iterative simulation trials that involve multi-agent systems designed to learn from their environment and their fellow agents.
The purpose of this project is to simulate social and behavioral evolution in a controlled environment and dynamic ruleset. Given a simple set of rules for how to interact with elements in an evironment–including other agents–can agents autonomously develop internal rules that are then shared with and learned by others?
This page is constantly under development.
- Iterative Learning for Rigid-Rule Actors. (07 August 2017). I’ve gone through the first iteration of my computational intelligence simulation system, and what I got was bunch of dumb actors moving around a giant grid and licking each other.
Research Writing Style
I’ve developed a style of reporting and writing about research that I conduct called IMEDDR, which stands for “Inspiration, Methods & Experimental Design, Discussion of Results.” In my PhD program, I became fed-up with the rigidity of the traditional IMRAD format for research write-ups, despite its time-honored nature and familiarity. I felt (and still feel) like IMRAD forces scientists to think about their work backwards, causing the research paper to flow in a disjunctive manner that is not reflective of the scientific thinking process that goes on as a reasearcher goes from inpsiration to experiment to discussion. the IMEDDR format is simple, concise, and follows my thought process as I go from what went into me wanting to experiment, what went into the actual experiment, and what became of the experiment.
I’d like the project to eventually examine the differences between developed rules for agents that grew up in highly stressful situations vs agents that grew up in relatively non-stressful situations.