January 2020
Mika Sumida — Network Revenue Management with Performance Guarantees: An Approximate Dynamic Programming Approach
Abstract: In network revenue management problems, we have a collection of resources with limited capacities to satisfy the requests for products arriving randomly over time. If we accept a request for a product, then we generate a product-specific revenue and consume the capacities of a combination of resources used by that product. The goal is to maximize the total expected revenue from the accepted product requests. Dynamic programming formulations to compute the optimal policy suffer from the curse of dimensionality. Instead, we provide an approximate policy with a performance guarantee. Our approximate…
Find out more »Somya Singhvi — Economically Motivated Adulteration in Farming Supply Chains
Economically motivated adulteration (EMA) is a serious threat to public health. In this paper, we develop a modeling framework to examine farms’ strategic adulteration behavior and the resulting EMA risk in farming supply chains. We study both “preemptive EMA,” where farms engage in adulteration to decrease the likelihood of producing low-quality output, and“reactive EMA,” where adulteration is done to increase the perceived quality of the output. We fully characterize the farms’equilibrium adulteration behavior in both types of EMA and analyze…
Find out more »IEOR/ORMS Undergraduate Pizza Welcome Party
Join us for pizza and fun next Wednesday, January 29th at La Val's Pizza. All IEOR/ORMS undergraduates are welcome. Come get to know your peers and learn more about the IEOR department, IISE, and APM.
Find out more »February 2020
Cem Randa – Structural Estimation of Kidney Transplant Candidates’ Quality of Life Scores
Abstract: This paper develops a framework for assessing the impact of changes to the deceased-donor kidney allocation policy taking into account the transplant candidates' (endogenous) organ acceptance behavior. To be specific, it advances a dynamic structural model of the transplant candidates' accept/reject decisions for organ offers. Our formulation models the national list (and its geographic structure) which is important for practical implementations, e.g.. for incorporating it in the Kidney Pancreas Simulation Allocation Model (KPSAM). Moreover, it allows various important features…
Find out more »Ikhlaq Sidhu — Can We Teach Students to Create, Innovate, and Lead?
Abstract: Skype, Zoom, Webex, Facebook voice chat and Whatsapp are all businesses built on VoIP, a disruptive technology that was controversial when Dr. Ikhlaq Sidhu developed and wrote its seminal patents while he was an engineer at 3com. In this talk, Ikhlaq will discuss his journey from lab engineer to industry innovator — and the concepts, mindset, and behaviors he discovered that are needed for innovation and commercial success. The talk includes methods for students at Berkeley to develop valuable…
Find out more »Lavanya Marla — Dynamic Disruption Management in Airline Networks under Airport Operating Uncertainty
Abstract: Operating disruptions result in enormous costs across spatial-temporal networks. For instance, air traffic disruptions result in flight delays, cancellations, passenger misconnections, creating high costs to aviation stakeholders. Businesses often design recovery plans in response to past disruptions while preparing for future disruptions. However, future disruptions often can only be characterized partially and probabilistically. We propose a joint stochastic reactive and proactive approach to disruption management (SRPDM), which optimizes airline disruption recovery given partial and probabilistic forecasts of future congestion…
Find out more »Berkeley–Columbia Meeting in Engineering and Statistics
The Berkeley–Columbia Meeting in Engineering and Statistics provides a biannual, interdisciplinary forum for research in Engineering, Finance, Mathematics and Statistics. The first meeting was held at UC Berkeley in 2016, and the second meeting was held at Columbia University in 2018. The third meeting will take place again on the UC Berkeley campus on February 28 and 29, 2020. Registration is not necessary. However, seating is limited—first come, first served. For complete schedule, speaker information, and details on how you…
Find out more »Jon Kleinberg — Fairness and Bias in Algorithmic Decision-Making
Abstract: As data science has broadened its scope in recent years, a number of domains have applied computational methods for classification and prediction to evaluate individuals in high-stakes settings. These developments have led to an active line of recent discussion in the public sphere about the consequences of algorithmic prediction for notions of fairness and equity. In part, this discussion has involved a basic tension between competing notions of what it means for such classifications to be fair to different…
Find out more »March 2020
Paul Grigas — Learning, Optimization, and Generalization in the Predict-then-Optimize Setting
Abstract: In the predict-then-optimize setting, the parameters of an optimization task are predicted based on contextual features and it is desirable to leverage the structure of the underlying optimization task when training a machine learning model. A natural loss function in this setting is based on considering the cost of the decisions induced by the predicted parameters, in contrast to standard measures of prediction error. While directly optimizing this loss function is computationally challenging, we propose the use of a…
Find out more »Sujay Sanghavi – Towards Model Agnostic Robustness [virtual talk]
Abstract: It is now common practice to try and solve machine learning problems by starting with a complex existing model or architecture, and fine-tuning/adapting it to the task at hand. However, outliers, errors or even just sloppiness in training data often lead to drastic drops in performance. We investigate a simple generic approach to correct for this, motivated by a classic statistical idea: trimmed loss. This advocates jointly (a) selecting which training samples to ignore, and (b) fitting a model…
Find out more »