Our mission at the Data Science Institute (DSI) is to enable excellence in data science research and applications across the Laboratory's core missions.

Data science has become an essential discipline paving the path of LLNL's key program areas, and the Laboratory is home to some of the largest, most unique, and most interesting data and supercomputers in the world. The DSI acts as the central hub for all data science activity—in areas of artificial intelligence, big-data analytics, computer vision, machine learning, predictive modeling, statistical inference, uncertainty quantification, and more—at LLNL working to help lead, build, and strengthen the data science workforce, research, and outreach to advance the state-of-the-art of our nation's data science capabilities. Read more about the DSI.

Data Scientist Spotlight

Marisol Gamboa

Marisol Gamboa

Computer Scientist

Marisol Gamboa thrives at the intersection of solving challenges in unique ways and mentoring the next generation. Over her 18-year career at LLNL, she has honed expertise in software engineering, web applications, and big data analytics by developing solutions for numerous defense and counterproliferation programs—such as tools that help Department of Defense personnel distill, combine, relate, manipulate, and access massive amounts of data in a timely manner. “The many lessons I’ve learned over the years have positioned me to tackle any challenge knowing that I am able to learn quickly and adjust to any situation in real-time,” she says. Gamboa is the Deputy Division Leader for LLNL’s Global Security Computing Applications Division as well as Computing’s Workforce Team Lead. She formerly co-directed the Data Science Summer Institute and created the annual Data Science Challenge with UC Merced. Active in outreach to young women and underrepresented minorities in STEM—including LLNL’s Women in Data Science regional events—Gamboa holds a B.S. in Computer Science from the University of New Mexico.


New Research in AI

Effective patient care mandates rapid, yet accurate, diagnosis. With the abundance of non-invasive diagnostic measurements and electronic health records (EHR), manual interpretation for differential diagnosis has become time-consuming and challenging. This scenario has led to wide-spread adoption of AI-powered tools,in pursuit of improving accuracy and efficiency of this process. Recently published research in Nature Scientific Reports introduces DDxNet, a multi-specialty diagnostic model for clinical time-series. The team found DDxNet to be highly effective across different modalities, diagnosis tasks, and data fidelity. The DDxNet model (GitHub repo) utilizes adaptively dilated causal convolutions coupled with dense connections for effective feature learning. The paper is co-authored by LLNL data scientist Jay Thiagarajan.

Five EEG abnormality graphs labeled as seizure, tumor present, healthy, eyes open, and eyes closed
Example EEG abnormality detection data from the different benchmark diagnosis problems considered in the paper. Note that only 1 of the 22 EEG channels is shown.

More AI news >>

 

Data Science Institute

 

Got data? We do. Public datasets are available via our Open Data Initiative.

coronavirus particle on a blue background

Browse our career opportunities. Or, have a better idea? Convince us! Send your resume and cover letter to datascience-jobs [at] llnl.gov.


We're growing the community. Check out our event schedule.


Dress for success. DSI gear is available now.

hoodie sweatshirt branded with DSI logo