deletions | additions       diff --git a/section_A_stochastic_optimization_model__.tex b/section_A_stochastic_optimization_model__.tex  index f6af31b..ff36163 100644  --- a/section_A_stochastic_optimization_model__.tex  +++ b/section_A_stochastic_optimization_model__.tex 
...
\section{A stochastic optimization model to reduce expected post-disaster response time through pre-disaster investment decisions.}  transportation network fortifying\\  Du and Peeta \cite{DuPeeta2014} develop the work of Peeta et al. \cite{PeetaEtal2010} in three directions: 1) assume that the link investment decisions are continuous rather than binary variables. 2) there is multiple failure scenarios for each link rather than just one scenario. 3) connectivity, traffic flow, and marginal survivability improvement determine link importance rather than just connectivity. In the study traffic flow is defined as expected number of times a link is used in the shortest path among nodes in all scenarios, and marginal survivability improvement indicates that links with low upgradation cost are more preferable candidates from an investment strategy perspective. The problem is then formulated as a two stage program solved via a heuristic algorithm to obtain feasible solution. The algorithm decomposes the model to a separate linear optimization model and an expected shortest path model,and iteratively solves them to converge a close-enough solution. The effectiveness of the model is evaluated on Sioux Fall City network data with 24 nodes and 76 links.