25 Years of INFORMS

 Anne Robinson, editor, 25 Years of INFORMS, EC2020 Volume 1.

The first section, By the Numbers, leverages Web of Science and Google Scholar for the most-cited articles in our journals over the past 25 years. The next section, Community Choice, represents articles selected by our current editors-in-chief as well as INFORMS Society/Fora leaders, and you’ll see the explanation of why these articles were selected as impactful content. Our third section – The Analytics Movement – highlights some of the critical drivers that led to the INFORMS community embracing analytics and the notion of descriptive, predictive, and prescriptive modeling as a way to describe our work at the proverbial cocktail party. Next, we capture some of the ways that INFORMS and its members have impacted society for the better, including highlights from the Edelman Award and Wagner Prize. Last, but certainly not least, is a collection of thoughts from INFORMS past presidents, some new material and some thoughts recorded during their tenure as INFORMS President.

INFORMS Analytics Collections Vol. 16: Advances in Integrating AI & O.R.

Ramayya Krishnan and Pascal Van Hentenryck, editors, Advances in Integrating AI & O.R., EC2021, Volume 16, April 19, 2021.

The INFORMS strategic initiative in AI resulted in a white paper that summarized the findings and provided a number of recommendations for the INFORMS community. This volume of Editor’s Cut complements the white paper and assembles a collection of papers from the INFORMS community that bridge the AI and O.R. communities. The papers are grouped into five categories:

1. Blending Predictive and Prescriptive Methods
2. AI/ML for Optimization Problems
3. Integrating Predictive and Causal Inference
4. Games, Control, Data-intensive Preference Estimation
5. Unstructured Data Analytics, AI and OR/MS – Innovative Applications

Smart "Predict, then Optimize"

Adam N. Elmachtoub and Paul Grigas, Smart "Predict, then Optimize", Management Science, 2021, 68(1):9-26. (Code in Julia, arXiv. 1st place, INFORMS Junior Faculty Interest Group (JFIG) Paper Competition, 2020)

語言焦慮和人才培育

吃飯時，走過某位老師的教室，聽到的都是英文。如前所述，為了國際生，系上的研究所課程使用英文教學。 其實，每一學期，至少有3門大學部的必修課，使用全英教學。

我有嚴重的語言焦慮 (language anxiety)，碰到英文越好的人，症狀越明顯。就算留學美國，也改不了基因。只好套用朋友的正向思考法，她在外商上班，心想上班免費學英文，不亦樂乎。

對系上的本地生而言，應該蠻辛苦的。但是，長期而言，不論是就業、旅遊、或者拓展人脈，應該有正向的助益。

另外一方面，經過全民數十年的努力，我們已經有知識輸出的能力。台灣科技人才短缺，如果能吸引周遭國家的人才，來台灣就學、就業，不僅可以補充白領階級的不足；在多元思維的環境下，也有助於國際化和推展國際貿易。 

昨天和學校長官開策略會議時，得知系上兩位博士班畢業生，正擔任菲律賓某大學的院長，具體說明， 這一種軟實力的擴散。

The Clinician and Dataset Shift in Artificial Intelligence

Samuel G. Finlayson et al., The Clinician and Dataset Shift in Artificial Intelligence, New England Journal of Medicine, 2021; 385:283-286.

A major driver of AI system malfunction is known as “dataset shift.” Most clinical AI systems today use machine learning, algorithms that leverage statistical methods to learn key patterns from clinical data. Dataset shift occurs when a machine-learning system underperforms because of a mismatch between the data set with which it was developed and the data on which it is deployed. For example, the University of Michigan Hospital implemented the widely used sepsis-alerting model developed by Epic Systems; in April 2020, the model had to be deactivated because of spurious alerting owing to changes in patients’ demographic characteristics associated with the coronavirus disease 2019 pandemic. This was a case in which dataset shift fundamentally altered the relationship between fevers and bacterial sepsis, leading the hospital’s clinical AI governing committee (which one of the authors of this letter chairs) to decommission its use. This is an extreme example; many causes of dataset shift are more subtle. In Table 1, we present common causes of dataset shift, which we group into changes in technology (e.g., software vendors), changes in population and setting (e.g., new demographics), and changes in behavior (e.g., new reimbursement incentives); the list is not meant to be exhaustive.

Deb Raji, There’s more to data than distributions, Mar 31, 2022. 

Jose G. Moreno-Torres et al., A unifying view on dataset shift in classification, Pattern Recognition, Volume 45, Issue 1, January 2012, Pages 521-530.