Theresa Mendoza edited untitled.tex  about 8 years ago

Commit id: dd99c93c423c9e068382d68fe50d466005f5f152

deletions | additions      

       

\subsection{Agent Based Models}  \label{sec:agentbasedmodels}  Agent based modeling is a powerful method to form simulations that models autonomous units to make decisions called agents. Agents create decisions like a human, has its own perspective and bound by certain decision rules but realistically also lacks the complete picture of the system. Agent based models can have rules and behaviors that could not have been predicted mathematically \cite{Ljubovic_2009)and one example in application to the research paper is the instance of traffic congestion.  Ljubovic et al. stated four instances where Agent Based Modeling is best used {:}  \begin{enumerate}  \item When behavior of individual units is non-linear, or it can only be described by a combination of if-then rules and thresholds.  \item When individual unit behavior includes memory, path, dependency//hysteresis, non-Markov behaviors or temporal dependencies such as adaptation and learning.  \item When interactions among agents are heterogeneous and can lead to network effects. With flow equations a homogeneous mixture is assumed, but due to the complex interaction topology such interactions can affect the whole system.  \item When averaging out eliminates important aspects of individual behavior that can affect the whole system.  \end{enumerate}  \section{Research Objectives}  \label{sec:researchobjectives}