A stochastic programming model with endogenous uncertainty for proactive supplier risk mitigation of low-volume-high-value manufacturers considering decision-dependent supplier performance

Reducing the potential damage caused by a wildfire is a problem of significant importance to land and fire managers. Fuel reduction treatment is a well-known method of reducing the risk of fire occurrence and spread on landscapes. However, officials seeking fuel reduction treatments on privately owned lands can only encourage it through incentive programs such as cost-share programs. This research developed a methodology that provides the basis for a decision-making tool to help managers allocate limited cost-share resources among a set of landowners to maximize wildfire risk reduction by implementing a hazardous fuel reduction treatment. A key feature of the methodology is that it incorporates uncertainty in the landowners’ decision of whether or not to implement treatment on their lands. The methodology is based on a stochastic programming model with endogenous uncertainty where the probability that a landowner accepts a cost-share offer to implement a fuel reduction treatment on their land depends on the offer amount. To estimate the probability that a landowner accepts a given cost-share offer amount, we used a predictive modeling technique to analyze landowner survey data. The results provide insight about the effects of different cost-share allocation strategies on the expected damage. Numerical experiments show that the risk-based allocation provides up to 37.3% more reduction in damage compared to other strategies that allocate equal cost-share amounts among landowners. Additionally, the results show that the solution quality is substantially sensitive to changes in the number of resource allocation levels.

Predicting Supplier Reliability in a Low Volume High Value Supply Chain Using Machine Learning

    •  Mike Sherwin, Medal, H., and C Mackenzie (2020). Predicting Supplier Reliability in a Low Volume High Value Supply

Chain Using Machine Learning. Tech. rep. Knoxville, TN: University of Tennessee.

Initial Orbit Selection For Forest Fire Monitoring with Descending Pass Observations

Hoskins, A. and Medal, H. (2020). Initial Orbit Selection For Forest Fire Monitoring with Descending Pass Observations.
Tech. rep. Knoxville, TN: University of Tennessee – Knoxville.

Risk-averse Bi-level Stochastic Network Interdiction Model for Cyber-security

Bhuiyan, T. H., Nandi, A. K., Medal, H., and M Halappanavar (2019). Risk-averse Bi-level Stochastic Network Interdiction
Model for Cyber-security. Tech. rep. Knoxville, TN: University of Tennessee.

Using DoDAF attack scenarios to develop Attacker-Defender Bi-level Mixed Integer Programming Models: A Wireless Network Jamming Application