Array Computing and the Evolution of Machine Learning

Travis E. Oliphant

January 7, 2020 at 2:00 P.M. in the Pearl Young Theater
(video within Langley firewall only)

Abstract

Computer Science must be regarded as a disciplined approach to connecting computational engines with data. The modern ability to store and share large amounts of data means computational performance scaling is now an inescapable part of a developer’s best practices. The history of array computing from the early days of APL to modern NumPy/Tensorflow/PyTorch—including what array/table computing means and how it should be more widely learned and applied—will be discussed by a pioneer in open-source computing. Dr. Oliphant will describe the emergence of popular Machine Learning and Artificial Intelligence engines that are available to all programmers as a result of wide-spread access to array computing and data-science tools. He will also offer potential solutions to the challenges that are leading to the current divergence of tools in the Python ecosystem for array computing.

Speaker

Dr. Travis E. Oliphant is a Founder and CEO/CTO of Quansight, an innovation incubation company that builds and connects companies with open-source communities to help both gain actionable, quantitative insight from their data. In 2019, Dr. Oliphant Co-founded OpenTeams, a partner company of Quansight, which aims to enable sustainable funding opportunities for open source software. Dr. Oliphant previously co-founded Anaconda Inc. and is still a Director. Since 1997, he has worked in the Python ecosystem, notably as the primary creator of the NumPy package and as a founding contributor of the SciPy package. Dr. Oliphant also started the Numba project and organized and led the teams that built Conda, Dask, Bokeh, and XND. Dr. Oliphant has a Ph.D. from the Mayo Clinic and B.S. and M.S. degrees in Mathematics and Electrical Engineering from Brigham Young University.