Home // SIMUL 2014, The Sixth International Conference on Advances in System Simulation // View article

Probabilistic Prognosis of Societal Political Violence by Stochastic Simulation

Authors:
André Brahmann
Uwe Chalupka
Hendrik Rothe
Torsten Albrecht

Keywords: Major Episodes of Political Violence; principal component analysis; classification; support vector machines

Abstract:
The current paper deals with the probabilistic prog-nosis of societal political violence levels of countries in the context of crisis prevention. The baseline is formed by two freely available datasets, whose relevance gets clarified in the first part of the paper. From these, a classification and a pre-diction modeling problem can be derived for which a Principal Component Analysis together with Support Vector Machines (SVMs) can be utilized as useful methods. Different SVM ker-nel functions have been investigated. To further perform the prediction step, a statistic modeling approach has been chosen that includes the computation of occurrence probabilities by stochastic simulation.

Pages: 172 to 177

Copyright: Copyright (c) IARIA, 2014

Publication date: October 12, 2014

Published in: conference

ISSN: 2308-4537

ISBN: 978-1-61208-371-1

Location: Nice, France

Dates: from October 12, 2014 to October 16, 2014