Date on Master's Thesis/Doctoral Dissertation

8-2021

Document Type

Master's Thesis

Degree Name

M.A.

Department

Anthropology

Degree Program

Anthropology, MA

Committee Chair

Jennings, Thomas

Committee Co-Chair (if applicable)

Smallwood, Ashley

Committee Member

Smallwood, Ashley

Committee Member

Gaughan, Andrea

Author's Keywords

GIS; archaeology; horticulture; Kentucky; settlement patterns; predictive modelling

Abstract

In this study, I explore the Late Archaic and Woodland settlement patterns (3,000 BC – 1,000 AD) in the Daniel Boone National Forest, Kentucky, and surrounding region within the context of the intensification of horticulture. GIS predictive modelling via automated learning algorithms are employed to explore various environmental variables that may have influenced where and why horticultural intensification occurred. Predictive models using random forest and maximum entropy are created and compared for the Late Archaic and Woodland periods. Results show only minimal variance between the Late Archaic and Woodland settlement patterns within the study area with slope and elevation identified as the most important environmental variables. Additional specificity and categorization of the data may serve to refine the findings and reveal further variances or similarities between the Late Archaic and Woodland periods.

Download

Share

COinS