Skip to main content
Expand AU Menu

Science

# Meet New Mathematics and Statistics Professor Michael Baron

Degrees

PhD statistics, University of Maryland
MS mathematics, St. Petersburg State University, Russia

Areas of Research

Sequential analysis and optimal sequential designs; change-point detection; multiple comparison; Bayesian Inference; application of statistics in epidemiology, clinical trials, semiconductor manufacturing, actuarial science, energy finance, and cyber security

What initially sparked your interest in mathematics/statistics?
Initially, it was the beauty of classical mathematics that I started learning in Russia from the world leaders of respective fields. It was the fantastic ability to prove very abstract results, seemingly unrelated to anything real, and then apply them to very down-to-earth practical situations. Working in statistics opens endless opportunities to derive mathematically optimal decision-making tools under uncertainty and then work with scientists and practitioners in diverse fields to solve modern cutting-edge problems, often including the key areas of national interest. Essentially, this is what I do for living.

What honed your interest to your specific areas of research?
My main passion is in a branch of statistics called sequential analysis. It covers all statistical methods designed for sequentially collected data, that is, in real time or online. A statistician observes the data as they arrive and decides:
(1) When it is best to stop collecting data and report results;
(2) What decision to take once sampling is seized.

I enjoy the dynamics of sequential designs, their flexibility, and application to many fields. For example, a multitude of clinical trials are conducted sequentially nowadays. This allows a trial to be stopped if a tested treatment is found to be inefficient or unsafe. On the other hand, it allows a study to be extendee until a definite result about its efficacy and safety is obtained. Optimizing the clinical trial, I can achieve the required accuracy at the minimal expected cost. This, in turn, means reduction of the cost of medicines, and ultimately, reduction in the cost of health care.

Another direction, still within sequential analysis, is change-point detection. A prompt identification of unexpected changes and unusual patterns allows us to prevent many unwanted significant deviations in industrial quality control, unusual traffic across the border, malicious computer applications, and other security threats. With my recent PhD student, we used these change-point methods to detect epidemic trends and predict influenza epidemics. Currently, with my collaborator in computer science, Professor Latifur Khan, we are developing applications in cybersecurity and discovery of biological threats. This research is supported by the National Science Foundation.

What brought you to AU?
A combination of many factors. The outstanding reputation of the university, distinguished faculty, and great opportunities for collaboration within AU as well as outside, especially federal agencies and research centers. I was also attracted by the interdisciplinary nature of the preferred candidate, according to the position announcement, and its Bayesian flavor. In Dallas, I was supposedly our department's Bayesian. And last but not least, I am a (proud) Maryland alumnus. So, I already have academic connections and good old friends in the area.

What are you hoping to accomplish at AU?
I will do my best to contribute to the growth and prosperity of AU and statistics within AU. I will use all my teaching experience to make my students motivated and inspired to learn statistical tools, especially for their chosen career paths. I will reach out to other schools at AU to establish research collaborations, probably starting with the School of International Service and Kogod School of Business. I will connect to federal funding agencies in the Washington, DC, area such as NSF, NSA, and NCI, work with their program directors, and serve on their panels. I will start working my ways to NIH funding and connections with the Department of Homeland Security and related agencies, including Defense Threat Reduction Agency (DTRA) and the National Geospatial Intelligence Agency (NGA). They co-support the NSF program "Algorithms for Threat Detection" that funds my current research on multichannel change-point detection. And also, I will learn to be the Washington Capitals fan now instead of a Dallas Stars fan.