Loading Events

« All Events

  • This event has passed.

National Labs Brown Bag Seminar Series: Dr. David Stracuzzi

February 21, 2019 @ 12:00 pm - 1:00 pm CST

| Free


Hosted by the National Security Collaboration Center

Faculty and students are invited to attend the Brown Bag Seminar series to listen to leading researchers from National Laboratories discuss their research in the Data Analysis, Artificial Intelligence/Machine Learning, and Cyber Security. You are welcome to bring your lunch and listen to this noon-time seminar series. It is also a great opportunity for faculty to investigate possible collaboration with National Labs investigators.

Dr. David Stracuzzi
Sandia National Laboratory

The role of uncertainty quantification in machine learning
Research Area: Machine Intelligence

February 21, 2019  || 12:00pm-1:00pm

Location:  North Paseo Building, NPB 4.120V

>> To register


Uncertainty is an inherent, yet often underappreciated, component of machine learning and statistical modeling.  Data-driven modeling often begins with noisy data from error-prone sensors collected under conditions for which no ground-truth can be ascertained.  Analysis then continues with modeling techniques that rely on a myriad of design decisions and tunable parameters.  The resulting models often provide demonstrably good performance, yet they illustrate just one of many plausible representations of the data – each of which may make somewhat different predictions on new data.

This talk provides an overview of recent, application-driven research at Sandia Labs that considers methods for (1) estimating the uncertainty in the predictions made by machine learning and statistical models, and (2) using the uncertainty information to improve both the model and downstream decision making. We begin by clarifying the data-driven uncertainty estimation task and identifying sources of uncertainty in machine learning.  We then present results from three different applications in both supervised and unsupervised settings.  Finally, we conclude with a summary of lessons learned and directions for future work.
Topics of Interest:

“My research broadly falls into machine learning and AI.  Specifically work that looks at:
– Automating complex tasks typically performed by people, but for various reasons, needs to be performed at CPU speed
-Analyzing, detecting, and extracting information from sensor data (including temporal data)
-Using sensor data to predict impending events based on a combination of background knowledge and historical data
-Incorporating domain expertise or theories into machine learning and statistical analyses

I’m also very interested in what we in 1400 have been calling “Human-Data Analytic Systems.”  This is the idea that if we are ever going to be successful in data sciences writ large, we’re going to have to figure out a more seamless integration of human analyst into the computational systems.  This is less about interfaces and more about (a) figuring out what the human analyst needs from the data, (b) figuring out how to incorporate what the human knows into the computational analysis, (c) making sure that we are computing the right information, and (d) making sure that we can present the results in such a way that the human interprets them correctly (especially if they are not a statistician by training).  There’s big cognitive sciences component to this in addition to machine learning and stats.”

**UTSA faculty, if you would like to meet with Dr. Stracuzzi individually on Feb. 21, please contact Patricia Geppert to set up a time.


February 21, 2019
12:00 pm - 1:00 pm


NPB 4.120V
1 UTSA Circle
San Antonio, TX 78217 United States
+ Google Map