Using Bayesian statistics for improving Agent based models and visa versa.
Agent based modelling;
Land use/land cover change;
Payment for ecosystem services;
Bayesian Network;
System Dynamics
GIS Developer
Agent Based Modeling (ABM), Agent Based Social System (ABSS), Multi-Agent Systems (MAS), Bayesian learning, Social networks Analysis (SNA), Socio ecological Dynamics.
Interested in IWRM approach, analyzing coupled human-water relationship, Hydrological modelling, Bayesian networks, Agent based modelling
I use agent-based systems, stochastic process, mass balance models and computational statistics in exploring human exposure assessment.
For my Ph.D. thesis, I developed a system to play poker.
I’m interested to see whether a similar approach can be applied to agent based models.
Didier’s Research:
are related to interoperability and conflation models in geospatial analysis and integrated modelling applications, particularly in the context of spatial data infrastructures such as GEOSS. This translates to a focus on geospatial statistics, geospatial patterns, outbreak detection and geospatial data mining in general, but also to data quality and uncertainty propagation principles in relation to geoworkflows connected to/using web services. Didier’s research centres on environmental agro-ecological geospatial models, and public health and spatial epidemiology applications. (see website)
My research interests include policy informatics and decision making, modeling in policy analysis and management decisions, public health management and policy, and the role of public value in policy development. I am particularly interested in less mainstream approaches to modeling that account for learning, feedback, and other systems dynamics. I include Bayesian inference, agent-based models, and behavioral assumptions in both my research and teaching.
In my dissertation research, I conceptualize state Medicaid programs as complex adaptive systems characterized by diverse actors, behaviors, relationships, and objectives. These systems reproduce themselves through both strategic and emergent mechanisms of program management. I focus on the mechanism by which citizens are sorted into or out of the system: program enrollment. Using Bayesian regression and agent-based models, I explore the role of administrative practices (such as presumptive eligibility and longer continuous eligibility periods) in increasing enrollment of eligible citizens into Medicaid programs.