Areas of Expertise (15)
Artifical Intelligence
Decision Analysis
AI Algorithms
Consumer Behavior
Data Analytics in Science and Medicine
Bayesian Methods
Decision Theory
Probability and Statistics
Statistical Analysis
Social Marketing
Data Analytics
Big Data
Data Security
Machine learning and Artificial Intelligence for Autonomous Systems
Data Analysis and Data Mining
Biography
James Scott, co-author of AIQ: How People and Machines Are Smarter Together, is a statistician and data scientist looking for better ways to solve problems that have frustrated industry professionals and researchers. His research interests include modern computational methods, Artificial Intelligence, and in Bayesian inference, including recent work on data-augmentation schemes for Bayesian computation; scalable algorithms; multiple testing and high-dimensional screening problems; prior choice in hierarchical models; and Bayesian methods in machine learning.
Scott is an associate professor in the department of information, risk, and operations management at the McCombs School of Business, The University of Texas at Austin. In 2013 he won the National Science Foundation CAREER award for his project, “Bringing Richly Structured Bayesian Models into the Discrete-Data Realm via New Data-Augmentation Theory and Algorithms."
"My goal is not to solve the marketing problem, or the finance problem, or the sentiment analysis problem," he says. "My goal is to look at all of those problems and see a common mathematical structure, some common principle that ties together a whole range of data sets and questions."
Scott received the Savage Award in 2010 for his dissertation on Bayesian statistics, “Bayesian Adjustment for Multiplicity.” The International Society for Bayesian Analysis presents the award annually to only two outstanding doctoral dissertations in the world.
His recent collaborative projects have involved applications in healthcare, security, and neuroscience. He has also done work in linguistics, political science, infectious disease, astronomy, and molecular biology.
Media
Documents:
Videos:
Audio/Podcasts:
Links (3)
Education (3)
Duke University: Ph.D., Statistics 2009
University of Cambridge: M.A.St., Mathematics 2005
The University of Texas at Austin: B.Sc., Mathematics and Plan II Honors 2004
Media Appearances (10)
Sweden actually protects its residents' data. America should take note.
Fortune online
2018-04-13
The ideas behind AI may be surrounded by a force field of technical jargon, but they’re surprisingly simple. How does AI work? Why does it depend so strongly on data? When and where does it go wrong? I promise you the answers are within your reach—and if you care about the world, few questions are more urgent today.
According to science, this is the most dangerous time to give birth
Yahoo Style UK online
2017-04-18
Speaking about the findings, Dr James Scott, an associate professor of statistics told The Sun: “There are all sorts of studies about the timing of deliveries, but what nobody had looked at before is whether there is some kind of proxy for how fatigued the doctors are.”
LIFE OR DEATH This is the most dangerous time to give birth for mum AND baby, experts warn
The Sun online
2017-04-17
Dr James Scott, an associate professor of statistics, said: “There are all sorts of studies about the timing of deliveries, but what nobody had looked at before is whether there is some kind of proxy for how fatigued the doctors are.
Birth risks rise at this point in a doctor’s day
Futurity online
2017-04-13
James Scott, an associate professor of statistics at the McCombs School of Business at the University of Texas at Austin, says that he and his fellow researchers hypothesized that the total number of hours worked during a shift may be a more important predicator of adverse outcomes than whether the delivery occurred during the weekend or in the middle of the night.
Latin American women report Zika virus alerts as reason for seeking abortion
"News Medical - Life Sciences & Medicine " online
2016-06-23
In November 2015, the Pan American Health Organization (PAHO) issued an epidemiological alert highlighting the risks of Zika. As more governments and health organizations began to respond, the researchers -- which also included co-authors James Scott from the University of Texas at Austin, Rebecca Gomperts and Marc Worrell from Women on Web, and Catherine Aiken from the University of Cambridge -- became interested in investigating the effects these health alerts had on women.
An Asset-Pricing Model for the Contagion Age: Polson and Scott
BloombergBusiness online
2011-12-07
The financial crisis and the meltdown in Europe have exposed the deficiencies of traditional asset-pricing models, particularly their inability to account for the effect of contagion from one market to another.
Human IQ and Artificial Intelligence Can Work Together, Business Professor Says
MSN Money
2018-06-22
Two professors, Nick Polson from the University of Chicago Booth School of Business and James Scott from the University of Texas at Austin, tried to put a face on the technology by writing a book that illustrates the beginning of AI through several examples of historical figures and other individuals who developed algorithms for humanity's different problems.
‘AIQ’ Review: Getting Smarter All the Time
Wall Street Journal online
2018-05-30
Artificial intelligence is a pervasive part of modern life—used to predict corn yields and map disease outbreaks, among other applications. Sam Kean reviews “AIQ” by Nick Polson and James Scott.
‘AIQ’ Review: Getting Smarter All the Time
Wall Street Journal online
2018-05-30
Artificial intelligence is a pervasive part of modern life—used to predict corn yields and map disease outbreaks, among other applications. Sam Kean reviews “AIQ” by Nick Polson and James Scott.
Human IQ and Artificial Intelligence Can Work Together, Business Professor Says
MSN Money online
2018-06-22
Two professors, Nick Polson from the University of Chicago Booth School of Business and James Scott from the University of Texas at Austin, tried to put a face on the technology by writing a book that illustrates the beginning of AI through several examples of historical figures and other individuals who developed algorithms for humanity's different problems.
Articles (10)
Multiscale Spatial Density Smoothing: An Application to Large-Scale Radiological Survey and Anomaly Detection
Journal of the American Statistical Association
2017-01-01
We consider the problem of estimating a spatially varying density function, motivated by problems that arise in large-scale radiological survey and anomaly detection.
Mixtures, Envelopes and Hierarchical Duality.
Journal of the Royal Statistical Society: Series B (Statistical Methodology)
2016-11-01
We develop a connection between mixture and envelope representations of objective functions that arise frequently in statistics. We refer to this connection using the term "hierarchical duality." Our results suggest an interesting and previously under-exploited relationship between marginalization and profiling, or equivalently between the Fenchel--Moreau theorem for convex functions and the Bernstein--Widder theorem for Laplace transforms.
Priors for Random Count Matrices Derived from a Family of Negative Binomial Processes
Journal of the American Statistical Association
2016-01-01
We define a family of probability distributions for random count matrices with a potentially unbounded number of rows and columns. The three distributions we consider are derived from the gamma-Poisson, gamma-negative binomial, and beta-negative binomial processes.
False Discovery Rate Regression: An Application to Neural Synchrony Detection in Primary Visual Cortex
Journal of the American Statistical Association
2015-01-01
Many approaches for multiple testing begin with the assumption that all tests in a given study should be combined into a global false-discovery-rate analysis. But this may be inappropriate for many of today's large-scale screening problems, where auxiliary information about each test is often available, and where a combined analysis can lead to poorly calibrated error rates within different subsets of the experiment. To address this issue, we introduce an approach called false-discovery-rate regression ...
Management of Fetal Malposition in the Second Stage of Labor: a Propensity Score Analysis
American Journal of Obstetrics and Gynecology
2015-03-01
Seeking to determine the factors associated with selection of rotational instrumental vs cesarean delivery to manage persistent fetal malposition, and to assess differences in adverse neonatal and maternal outcomes following delivery by rotational instruments vs cesarean delivery.
The Bayesian Bridge
Journal of the Royal Statistical Society: Series B (Statistical Methodology)
2014-01-01
We propose the Bayesian bridge estimator for regularized regression and classification. Two key mixture representations for the Bayesian bridge model are developed: (1) a scale mixture of normals with respect to an alpha-stable random variable; and (2) a mixture of Bartlett--Fejer kernels (or triangle densities) with respect to a two-component mixture of gamma random variables.
Nonparametric Bayesian Testing for Monotonicity
Biometrika
2013-04-11
This paper studies the problem of testing whether a function is monotone from a nonparametric Bayesian perspective. Two new families of tests are constructed.
Bayes and Empirical-Bayes Multiplicity Adjustment in the Variable-Selection Problem
The Annals of Statistics
2009-12-31
This paper studies the multiplicity-correction effect of standard Bayesian variable-selection priors in linear regression.
Feature-Inclusion Stochastic Search for Gaussian Graphical Models
Journal of Computational and Graphical Statistics
2007-12-31
We describe a serial algorithm called feature-inclusion stochastic search, or FINCS, that uses online estimates of edge-inclusion probabilities to guide Bayesian model determination in Gaussian graphical models.
Social