Tomorrow (March 11th, 2016)  I am giving a 20 minute presentation at the Northeast Archaeological Predictive Modeling Symposium, hosted by the New Hampshire Department of Transportation and the New Hampshire Division of Historical Resources.  My talk covers some of the broad lessons learned from creating the statewide Pennsylvania Archaeological Predictive Model Set.  This post is a place to share the presentation, link to download the presentation, link to download the reports (Volume 7 has the most concise wrap-up), and associated information.

As with most presentations, 20 minutes can barely scratch the surface of the advances that the PA Model made or the implications it has on the broader practice of archaeological predictive modeling, but it is a start.  Further, unfortunately the slides are presented as only graphics without the context of the symposium or my talk.  However, this should give some information to dwell on.  Finally, the animated gif of fits based on background samples (slide 14) is not animated in the PDF, but I posted about that very topic here.

Abstract

A recently completed archaeological predictive model (APM) of the state of Pennsylvania provided an unprecedented opportunity to explore the current status of APM methods and extend them based on current methods derived from related scientific fields, medicine, and statistical computing.  Through this process many different types of models were created and tested for validity, predictive performance, and adherence to archaeological theory.  One result of this project is a comprehensive view of the problems that beset existing APM methodologies, solutions to some of these problems, and the nature of challenges that we will face going forward with new techniques.  Most, if not all of the findings of this project are applicable to the eastern deciduous United States, and much of the methodological scope is useful to APMs in any geography.   This paper will discuss the primary lessons learned from this project in regards to archaeological data, modeling methods, and theory, as well as touch on best-practices for future APM efforts.

~ Matthew D. Harris, AECOM – Burlington, NJ

Presentation

Notes:

• I will do a series of posts on the PA Model to get some more details out into the wild.
• I am not crazy about using LinkedIn’s slideshare.net, but it conveniently integrated with WordPress; so be it.
• There are lots of hand-wavy treatments in this talk (and slides).  I would like to spend another 20 minutes talking about spatially explicit clustered K-folds CV strategies or Bayes inference, but we archaeologists have a lot of ground to cover and the basics of a more rigorous approach for identifying potential archaeological site locations needs to be discussed and agreed upon more widely across the field.  Until then, the in-the-weeds technical stuff only adds to a rather small echo chamber.
• If you have any questions about this topic, please ask!