Likelihood Estimation of Intentional Events in Risk Management: Evidence Based Intelligent Morphological Analysis Approach


Akgün İ.

Intelligent Techniques in Engineering Management, Cengiz Kahraman, Sezi Çevik Onar , Editör, Springer-Verlag , Aarau, ss.45-75, 2015

  • Yayın Türü: Kitapta Bölüm / Araştırma Kitabı
  • Basım Tarihi: 2015
  • Yayınevi: Springer-Verlag
  • Basıldığı Şehir: Aarau
  • Sayfa Sayıları: ss.45-75
  • Editörler: Cengiz Kahraman, Sezi Çevik Onar , Editör
  • Marmara Üniversitesi Adresli: Hayır

Özet

Risk management as a special Engineering Management field involves understanding the risks called risk assessment. Risk assessment is a highly complex strategic activity and requires a structured quantified methodology for identifying undesired events and estimating their likelihoods . However, undesired events like human caused intentional events are deliberate, innovative, and unpredictable acts unlike accidental failures and different from random events. Although there is a significant increase in the number of intentional events such as homeland and cyber security events in recent years, current risk assessment methods are not suitable for risk assessment of intentional events. In this study, a novel intelligent technique called evidence based Morphological Analysis (EMA) model based on Dempster-Shafer theory of evidence (DST) and Morphological Analysis (MA) methodology is proposed to quantify the likelihood of intentional events as threats by identifying them. It is based on the appropriate uncertainty model and supplementary methods for a risk management of a critical facility protection considering both information at hand and information requirements of decision makers. The proposed approach is presented step by step and applied to a simple case study on an airport. The two common DST combination rules are also analysed by considering decision maker’s attitude either risk averse or risk seeking. The results show that EMA analyses a wide range of plausible threat scenarios more easily than hierarchical techniques as tree structures with modest computational effort. Therefore, EMA can be used to reason about risk assessment by providing required output data precision for comparing and ranking of scenarios systematically.