As a Department of Energy National Laboratory, Berkeley Lab hosts many publicly available scientific datasets of use in machine learning applications.
As a Department of Energy National Laboratory, Berkeley Lab hosts many publicly available scientific datasets of use in machine learning applications.
As a Department of Energy National Laboratory, Berkeley Lab hosts many publicly available scientific datasets of use in machine learning applications.
Hosted by Berkeley Lab's Center for Advanced Mathematics for Energy Research Applications (CAMERA) this three-day series of talks, tutorials, and discussions will share recent developments in using AI/ML methods to enable autonomy and accelerate discovery. All researchers, including students and other early-career scientists, are welcome. Learn more.
Hosted by Computing Sciences at Berkeley Lab, this school brings together researchers and engineers for lectures and tutorials on state-of-the-art deep learning methods and best practices for running deep learning on high performance computing systems. Register to attend weekly live talks. Can't make it? View recordings of past lectures on YouTube. Learn more.
This half-day virtual workshop for Berkeley Lab researchers featured speakers from across the lab, industry, and academia discussing ongoing work at the interface of (meta)genomic data and deep learning.
The Computing Sciences Area at Berkeley Lab and the Association for High Speed Computing hosted the inaugural of this annual, invitation-only meeting. This year's discussions centered on deep learning. Learn more.
Hosted by Computing Sciences at Berkeley Lab, this school brought together researchers and engineers for lectures and tutorials on state-of-the-art deep learning methods and best practices for running deep learning on high performance computing systems. Video recordings of the sessions and presentations are available.
The event featured results from advanced data analytics and machine learning projects on the NERSC Cori system. And ongoing projects being led by NERSC, Intel, Cray, and five Intel® Parallel Computing Centers were featured in the talks. Presentations are available for download.
At this one-day summit, a variety of automotive, semiconductor, and artificial intelligence experts from industry, academia, and the national laboratories explored advanced microelectronics and computing approaches to help meet future energy, cost, and computational requirements for connected and automated vehicles (CAVs). The event was co-hosted with Sandia National Laboratories. An agenda is available online.
Berkeley Lab's Energy Technologies Area and Computing Sciences Area hosted this workshop to bring together researchers from throughout the Berkeley ecosystem with common interests in transportation, infrastructure, the grid, artificial intelligence, and materials science. More than 75 attendees, including select representatives from industry, UC Berkeley, lab leadership, all divisions of ETA, Computing Sciences, and Earth and Environmental Sciences, participated in collaborative discussions on cybersecurity threat detection; intelligence opportunities in prediction, artificial intelligence in materials discovery; and integrated modeling of transportations, buildings and the grid.
This invitation-only event brought together early and mid-career applied mathematicians to explore current and future challenges in machine learning and data analytics. Sponsored by the white paper that lays out a vision for the next 10-20 years for the labs. An agenda and keynote speakers' slides can be found on the workshop's website.
NERSC staff delivered a full-day tutorial on Deep Learning at Scale covering a working knowledge of deep learning on HPC class systems including core concepts, scientific applications and techniques for scaling. Attendees were provided NERSC training accounts and example Jupyter notebook-based exercises as well as datasets for hands-on experimentation with ML concepts such as training, inference and scaling of deep neural network machine learning models. An agenda is available on the SC18 website.
Data Day, now in its third year, is a data-centric event that brings together researchers who use, or are interested in using, NERSC systems for data-intensive research. The event features the latest data-focused tools for scientific computing, training sessions on machine learning, Python and more, along with presentations from scientists already using data tools and services in their work plus a hack-a-thon and tutorials.
The inaugural workshop on machine learning for science at Berkeley Lab, ML4Sci, was held jointly with the annual NERSC Data Day and featured scientific machine learning applications at the lab in high energy physics, nuclear physics, cosmology, chemistry, biosciences, materials engineering, climate and high performance computing. There were overviews of ML methodology and technology and hands-on training for deploying ML applications on NERSC platforms. A full agenda and slides are available on the 2018 ML4Sci workshop website.
The Big Data Summit 2018 featured results from advanced data analytics and management projects on the NERSC Cori system. The summit featured talks about ongoing projects being led by NERSC, Intel, Cray, and five Intel Parallel Computing Centers (represented by Oxford University, University of Liverpool, University of California Berkeley, University of California Davis, and New York University). Recordings of the presentations are available on the summit's web page.
Berkeley Lab’s growing involvement in deep learning research and development resulted in staff members presenting papers and posters for the first time at the 2017 Conference on Neural Information Processing Systems (NIPS). NIPS is a machine learning and computational neuroscience conference that includes invited talks, demonstrations and oral and poster presentations of refereed papers. Read More.