BEGIN:VCALENDAR VERSION:2.0 PRODID:Linklings LLC BEGIN:VTIMEZONE TZID:America/New_York X-LIC-LOCATION:America/New_York BEGIN:DAYLIGHT TZOFFSETFROM:-0500 TZOFFSETTO:-0400 TZNAME:EDT DTSTART:19700308T020000 RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU END:DAYLIGHT BEGIN:STANDARD TZOFFSETFROM:-0400 TZOFFSETTO:-0500 TZNAME:EST DTSTART:19701101T020000 RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU END:STANDARD END:VTIMEZONE BEGIN:VEVENT DTSTAMP:20210402T160039Z LOCATION:Track 1 DTSTART;TZID=America/New_York:20201117T130000 DTEND;TZID=America/New_York:20201117T134500 UID:submissions.supercomputing.org_SC20_sess300_inv114@linklings.com SUMMARY:Trustworthy Computational Evidence Through Transparency and Reprod ucibility DESCRIPTION:Invited Talk\n\nTrustworthy Computational Evidence Through Tra nsparency and Reproducibility\n\nBarba\n\nMany high-performance computing applications are of high consequence to society. Global climate modeling i s an historic example of this. In 2020, the societal issue of greatest con cern, the still-raging COVID-19 pandemic, saw a legion of computational sc ientists turn their endeavors to new research projects in this direction. Applications of such high consequence highlight the requirement of buildin g trustworthy computational models. Emphasizing transparency and reproduci bility have helped us build more trust in computational findings. In the c ontext of supercomputing, however, we may ask: how do we trust results fro m computations that cannot be repeated? Access to supercomputers is limite d, allocations are finite and machines are decommissioned after a few year s. I had the distinction to serve as SC19 Reproducibility Chair, and contr ibute to the strengthening of this initiative for SC. I was also a member of the National Academies study Committee on Replicability and Reproducibi lity in Science, which released its report last year. There, reproducibili ty is defined as "obtaining consistent computational results using the sam e input data, computational steps, methods, code and conditions of analysi s." We should ask how this can be ensured, certified even, without exercis ing the original digital artifacts. This is often the situation in HPC. It is compounded now with greater adoption of machine learning techniques, w hich can be opaque. The ACM in 2017 issued the Statement on Algorithmic Tr ansparency and Accountability, targeting algorithmic decision-making using data models. Among its seven principles, it calls for data provenance, au ditability, validation and testing. These principles can be applied not on ly to data models, but to HPC in general. In this talk, I want to discuss the next steps for reproducibility: how we may adapt our practice to achie ve what I call unimpeachable provenance, and achieve full auditab ility and accountability of scientific evidence produced via computation.\ n\nTag: Applications, Computational Science, Extreme Scale Computing, Mach ine Learning, Deep Learning and Artificial Intelligence\n\nRegistration Ca tegory: Tech Program Reg Pass, Exhibits Reg Pass END:VEVENT END:VCALENDAR