Seminar Series
Wednesday, 9/18 - Ubiquitous Machine Learning: Deep Data Representation for Next Generation Systems
Dr. Eric Larson, Associate Professor in Computer Science in the Bobby B. Lyle School of Engineering
This talk will overview data science related research topics in Dr. Larson’s lab. These research topics explore the interdisciplinary relationship of machine learning and signal/image processing with the fields of security, mobile health, education, psycho-visual psychology, human-computer interaction, and ubiquitous computing. Emphasis will be given to deep data representations that support cyber-security, education, and healthcare via the integration of machine learning and sensing. These include (1) using machine learning in robotic surgery assessment, (2) assessing oral reading fluency and vocabulary acquisition, and (3) investigate information leakage in pervasive and mobile devices.
Moody Hall Classroom 126 | 2:00 pm - 3:00 pm
Slides from Dr. Larson's presentation can be found here.
Wednesday, 10/23 - Rebuilding Earthquake Catalogs Using Machine Learning
Dr. Heather R. DeShon, Professor of Geophysics; Chair, Huffington Department of Earth Sciences
Machine learning (ML) phase detection and association has altered the landscape for the rapid creation of earthquake catalogs. Many rapid deployments in urban areas, such as North Texas, and in offshore environments create scenarios the push the limits of available training datasets. Here, Dr. DeShon will overview some of the applications of ML techniques used to develop earthquake catalogs at SMU and review the scientific insights that these catalogs provide on seismogenic processes in oil and gas basins and subduction zone environments.
Moody Hall Classroom 126 | 2:00 pm - 3:00 pm
Monday, 11/23 - SmartCADD: Artificial Intelligence: Quantum Chemistry Empowered Drug Discovery Platform With Explainability
Dr. Elfi Kraka, Professor and Harold Jeskey Chair, Department of Chemistry
Dr. Corey Clark, J. Lindsay Embrey Professor; Deputy Director, Research, SMU Guildhall; Assistant Professor, Department of Computer Science
The discovery of a novel drug requires approximately 12 years and around 1 billion dollars. There is the great expectation that computer aided drug design (CADD) can help to shorten this process. However, CADD requires a complex machinery ranging from screening billions of drug-like molecules as potential candidates (microscale) to high-accuracy calculations of target-lead properties at the quantum chemical (QM) level for the most promising drug-candidates (microscale). We will elucidate in this talk how SmartCADD combines the macroscopic and microscopic ends, leading for the first time to a holistic picture of the whole drug design process. Dr. Kraka will provide insights on the QM aspects of the platform and Dr. Clark will focus on the specific AI tools utilized in SmartCADD.
Moody Hall Classroom 126 | 2:00 pm - 3:00 pm
Slides from Drs. Kraka and Clark's presentation can be found here.