China
Africa
Explore chapters and articles related to this topic
Basic principles of ERM
Published in Terje Aven, Shital Thekdi, Enterprise Risk Management, 2019
In the following we will present and discuss various features of this framework: Overall modelRisk management in different stages of a projectUsing and controlling assumptionsClassifications of threats and risksDecision-making, choice of alternatives and measuresDiscussionConclusions
Robust Quality for Analytics
Published in Rajesh Jugulum, Robust Quality, 2018
It is important for organizations to develop model performance risk management framework by using evaluation indices like RQI, and such a framework should be integrated into the organization’s overall risk management policy. In addition, the model risk management framework should periodically be updated and evaluated by involving an independent team of experts. Model performance risk management will also help to evaluate the effectiveness of using analytical models including artificial intelligence and machine learning techniques and the investments made on them.
Applications of Machine Learning in Industrial Sectors
Published in Pedro Larrañaga, David Atienza, Javier Diaz-Rozo, Alberto Ogbechie, Carlos Puerto-Santana, Concha Bielza, Industrial Applications of Machine Learning, 2019
Pedro Larrañaga, David Atienza, Javier Diaz-Rozo, Alberto Ogbechie, Carlos Puerto-Santana, Concha Bielza
Financial institutions have used machine learning techniques for a number of operational (or back-office) applications. These applications include capital optimization by banks, model risk management and market impact analysis.
Risk-based pavement maintenance planning considering budget and pavement deterioration uncertainty
Published in Structure and Infrastructure Engineering, 2022
Amirhossein Fani, Amir Golroo, Hamed Naseri, S. Ali Mirhassani, Amir H. Gandomi
Considering risk in the pavement M&R optimization problem is vital as a significant amount of money is generally spent on pavement networks. In the pavement management system, the risk is generally defined as increasing the total cost (Damnjanovic & Zhang, 2008; Seyedshohadaie et al., 2010). Therefore, in the current study, risk was defined as the total cost of uncertainty scenarios with the highest M&R costs. That is, this study aims to minimize the total cost of the most expensive uncertainty scenarios to reduce the risk if these scenarios happen. Conditional Value-at-Risk (CVaR) was employed as an indicator to model risk in the pavement M&R optimization problem. Risk can be defined as investment loss values (i.e. total M&R cost in the current study), and Value-at-Risk (VaR) is defined based on this concept. VaR investigates the loss probability and calculates the most expected loss in the planning horizon at the confidence level of In other words, VaR gives the maximum possible loss with a specified confidence level (Christoffersen, 2009). CVaR calculates the average value of the loss which exceeds the VaR value (Rockafellar & Uryasev, 2000). CVaR accounts for losses exceeding VaR. Figure 2 illustrates CVaR, VaR, and the most expected loss. VaR represents the maximum loss with the probability of and can be calculated using (Roustai, Rayati, Sheikhi, & Ranjbar, 2018):
A newsvendor model with autocorrelated demand under a time-consistent dynamic CVaR measure
Published in IISE Transactions, 2019
Ye Shi, Layth C. Alwan, Christopher Tang, Xiaohang Yue
An often-mentioned way in which to model risk aversion is using Value at Risk (VaR). Specifically, for any given confidence level , the VaR of an uncertain “loss” random variable X with probability distribution is defined as , where measures the probability that the loss of more than z is less than η so that where η is the degree of risk aversion; hence, smaller η is associated with greater risk aversion. In the context of an optimization problem, previous studies typically find that it is optimal to order less when the newsvendor becomes more risk averse (i.e., as η decreases).