Virtual Graduate Student Colloquiums

All dates for this event occur in the past.

Zoom Meeting ID: 966 9446 2785

Password: 071928

 

Presenter: Abdullah Alomair

Advisor: Ted Allen

Committee members: Mike Groeber, Antonio Conejo and Rajiv Ramnath

Cyber security is one of the top domestic and international security threats. In this research, we focus on experimental design and analysis for adversarial environments, resilience engineering, and cybersecurity education/ Therefore, we consider three issues: (1) How to analyze single and two period gaming experiments to permit Nash and correlated equilibria enumeration efficiently? (2) How to simulate cyber-attack and defense including within gaming exercises to support strategy? (3) How to design a resilient system (attack preparation and recovery) for adversarial attacks that optimize the resources allocation to countermeasures and gain the ability to restore steady state baseline. We seek to analyze the off-line experiments with multiple factors (variables), considering unknown game parameters. We propose a resilience engineering system that helps IT administration and decision makers to optimize the resource allocation among the available options of two separated stages: preparedness (for the prevention) and recovery (for mitigation). The preparedness robust design problem is modeled as a two-stage stochastic program with recourse. Startup and sequential procedures for efficient estimation to support robust design and proposed and analyzed.


Presenter: Arthur Yang

Advisor: Guzin Bayraksan

Committee members: Antonio Conejo and Cathy Xia

Description: Distributionally robust multistage stochastic programs that involve mixed-integer decision variables can be used to model many real-world applications. However, they are one of the most challenging optimization models to solve. There are three main topics of this research. First, a multi-objective multistage stochastic mixed-integer quadratic mathematical model for water allocation with hedging rule is proposed and investigated. This model is applied to the multi-reservoir system located at Taize River Basin in China. Second, a Distributionally Robust Optimization (DRO) version of this type of problem is considered. To handle the ambiguity of the true distribution, a set of possible distributions are defined using ϕ-divergences. Then, computationally efficient approximation and bounding techniques for this class of problems are explored. Finally, a solution algorithm is investigated for DRO formed via ϕ-divergences that involves mixed-integer decision variables. Unlike previous studies, this proposed algorithm does not require any structural assumptions.


Presenter: Wenchen Zhou

Advisor: Allen Yi

Committee members:  Jose Castro and Thomas Raasch

This research will develop a novel process of fabricating freeform nickel molds used in precision glass modeling process, extend the applications of freeform surfaces in the imaging field by developing an infrared compound eye, and propose a new freeform metrology method based on SHS.  Overall, this research seeks improvements in the fabrication, the application and metrology of freeform optics.


Presenter: Abolfazl Zolfaghari Abbasghaleh

Advisor: Allen Yi

Committee members: Jose Castro and Michael Groeber

Fabrication techniques of precise optical components have been improved recently. However, except molding-based methods, they are not appropriate for mass production. Molding-based methods are suitable for high production rate, but these techniques are less capable to fabricate complex precise optical components properly due to restrictions in mold making processes. To deal with these limitations, a process based on nickel electroforming has been developed to replicate mold inserts directly from a plastic optical component, which itself could be fabricated by any fabrication methods with high flexibility. Then, using the mold insert and precision compression molding, copies of the optical component are produced. Feasibility of the proposed fabrication technique has been demonstrated through replication of glass, infrared and plastic microlens array. Moreover, the fabricated microlens arrays have been evaluated in terms of accuracy, surface quality and optical performance. Based on the obtained results, the developed manufacturing method is a promising method to fabricate wide range of optics.