Test pages

From GeoMod

[[1]Project Home]

Outline:

The Upper Palaeolithic, Old Stone Age, was a period in human development characterized by the development of stone tool technology. Chert, a family of crypto-crystalline sedimentary minerals, was the primary parent material in the manufacture of stone tools. During this period, Aurignation hunter-gather groups lived in the current Aquitaine region of France. Lithic artifacts of Aurignation groups have been collected as assemblages of toolstones. Outcrops of chert are present in the proximity, of assemblage sites. The spatial relationships between chert outcrops and assemblages, are products of the reasoning process of Aurignation hunter-gatherers. By modeling these relationships it is possible to derive characteristics of the spatial reasoning processes of Aurignation hunter-gatherers to the paleolithic landscape. Spatial reasoning can be characterized as a set of Beliefs, Desires, and Intentions(BDI), has been used as operational framework for random agent models. This work details the development of a random agent model of a Aurignation hunter-gatherer group called Chertman.

Through BDI, Chertman models the collection of chert from the surrounding environment to produce a simulated assemblage of material that can be directly compared to observed assemblages by mass and composition. Presented first is a literature review of the development and recent trends in Agent Based Modeling(ABM), pedestrian archeology, and Sensitivity Analysis(SA) upon which the methodology and initial model input parameters of this project are based. Second, model input parameters are screened for their fundamental contribution to model output, by the One at A Time(OAT) method (Morris,1991). Model input parameters are then refined in both character and resolution to redefine the model's sample space. Model output is then validated and optimized by simulation annealing, a type of genetic algorithm, operating through the covariance of samples from multiple runs of model output to observed assemblage data. Analysis of model output is performed to construct a (BDI) framework of Aurignation spatial reasoning. Emergent patterns of the random model are characterized. Finally, recommendations, for further research are made.

Title

ii Copyright Notice ABM

iii Dedication ABM

iv Acknowledgments ABM

1. Abstract ABM
2. Preface ABM

1. List of Tables ABM
2. List of Figures ABM
3. Key to Symbols and Abbreviations

1. Introduction
   1. Scope of the Study
   2. Description of the Problem
   3. Purpose of the Study
   4. Limitations of the Study
2. Prior Research
   1. Literature Review Agent Based Modeling
   2. Literature Review Pedestrian Archeology
   3. Literature Review Sensitivity Analysis
   4. Literature Review Optimization
3. Sensitivity Analysis/Simulation Annealing
   1. Model Agent Properties
   2. Chertman Code
   3. Simulation ABM
4. Conclusions ABM
5. Recommendations ABM
6. Bibliography/References
   1. Glossary ABM
   2. Bibliography ABM
   3. Agent based modeling resources
   4. Appendix ABM