Authors
Erman Cakit, Waldemar Karwowski
Publication date
2015
Journal
International Journal of Machine Learning and Computing
Volume
5
Issue
3
Pages
252-257
Description
The main purpose of this study was to investigate the relationship between adverse events and infrastructure development investments by estimating the number of adverse events in an active war theater based on the allocation of infrastructure development projects utilizing a fuzzy inference systems (FIS) approach. The considered model input variables included the total number of economic improvement projects and their associated budgets at different time periods in Afghanistan between 2002 and 2009. The output variables were the estimated numbers of people killed, wounded, and hijacked in different sectors of Afghanistan in 2010. Six different prediction models were developed and tested with an independent datasets. The prediction accuracy of each FIS model was evaluated and compared based on the mean absolute errors (MAE). It was concluded that the FIS is a useful modeling approach that can be applied under scenario-based conditions to support decision makers in analyzing historical economic data on how allocation of regional infrastructure development funds can best help reducing the onset of adverse events in an active war theater.
Total citations
201520162017201820192020212241