Category Archives: Events

Seminar on Machine Learning Explainability

Dragutin Petkovic from the Computer Science Department at San Francisco State University presented “Toward Explainable Machine Learning – RFEX: Improving Random Forest Explainability” at last week’s Mobilize Center seminar. The abstract for the talk is below and the presentation slides are available here.
 
Abstract: Machine Learning (ML) methods are now influencing major decisions about patient care, new medical methods, drug development and their use and importance are rapidly increasing in all areas.  However, these ML methods are inherently complex and often difficult to understand and explain resulting in barriers to their adoption and validation. We define explainability in ML as easy to use information explaining why and how the ML approach made its decisions. We believe that much greater effort is needed to address the issue of ML explainability because of the ever increasing use and dependence on ML in many applications and the need for increased adoption by non-ML experts. 
 
In our talk, we will 1) summarize a workshop discussion on ML explainability organized jointly with Profs. L. Kobzik and C. Re at the 2018 Pacific Symposium on Biocomputing (PSB) and 2) describe our work on Random Forest Explainability (RFEX) (joint work with Prof. R. Altman, M. Wong and A. Vigil). RFEX provides easy-to-interpret explainability summary reports from trained RF classifiers to improve the explainability for users who are often non-experts. We tested RFEX with the FEATURE program to predict functional sites in 3-D molecules based on their electrochemical signatures (features). Through formal usability testing with expert and non-expert users, we found the RFEX explainability report significantly increased explainability and user confidence in RF classification.

Young investigators competition at ISMB 2018 -Deadline is April 5th, 2018

A competition for young investigators associated with the Big Data to Knowledge (BD2K) Initiative will be held in conjunction with the 26th Conference on Intelligent Systems for Molecular Biology (ISMB 2018). ISMB 2018 will be held on July 6-10 in Chicago, Illinois. The conference is sponsored by the International Society for Computational Biology.

ISMB 2018 is expected to include several sessions featuring talks from officials of the National Institutes of Health on biomedical data science, projects under the Big Data-to-Knowledge (BD2K) Initiative and training programs. The conference brings together scientists from computer science, molecular biology, mathematics, statistics and related fields. Its principal focus is on the development and application of advanced computational methods for biological problems.

As part of BD2K, there will be a competition for Young Investigators. 6 individuals will be selected for oral presentation and travel awards, including ISMB registration and $500 towards your travel support to ISMB 2018. These individuals will be invited to give talks during a special NIH BD2K session at ISMB, and the abstracts will be recognized at the ISMB award ceremony.

To be considered for this award:

• Submit an abstract to ISMB 2018 here: https://www.iscb.org/ismb2018-submit/ismb2018-abstracts
• Send a copy (PDF version) of this abstract and submission to deniseluna@bd2kccc.org by April 5th, 2018 deadline. When submitting, please include (1) your affiliated BD2K project when you submit this to us, including BD2K grant number, and (2) a letter of support from the project PI or training director. (If you are the project PI, no need for this letter).
• Awardees will be informed by May 1st, 2018.

Eligibility: to qualify for this award, a Young Investigator is any individual who is (1) an existing graduate student, or (2) a postdoctoral fellow in a BD2K-funded laboratory/program; or (3) any individuals who are considered an NIH Early Stage Investigator (ESI) and have received funding support from an NIH BD2K Award (the ESI does not have to be a PI on the NIH award). Two or more finalists will be selected from each of these three categories.

There will also be a networking event for Young Investigators at the evening of Sunday July 8th, 2018, with more details to follow. If you are interested in participating in this event, please email: deniseluna@bd2kccc.org.

Outstanding Paper Award for Surgical Science from the Spine Journal

Mobilize Center faculty Matthew Smuck and colleagues won the 2017 “Outstanding Paper: Surgical Science” at the North American Spine Association (NASS) 32nd Annual Meeting for their paper“Objective Measurement of Function following Lumbar Spinal Stenosis Decompression Reveals Improved Functional Capacity with Stagnant Real-Life Physical Activity” in The Spine Journal. Results showed improvements in self-reported function and objectively measured physical capacity 6 months post-surgery, but not in physical performance as measured by continuous activity monitoring. Read more.

NIPS “Learning to Run” Challenge Winners Announced

NNAISENSE, an artificial intelligence company based in Switzerland, beat out 441 other participants to win first prize in the “Learning to Run” competition co-organized by researchers from UC Berkeley, EPFL, and the Mobilize Center. The competition was part of the Neural Information Processing Systems (NIPS) 2017 Competition track. In the competition, participants were tasked with developing a controller to enable a physiologically-based human model to navigate a complex obstacle course as quickly as possible. NNAISENSE’s model ran at 4.60 m/s. The second prize winner’s model (Beijing University, China) ran at 4.17 m/s, and the third prize winner’s model (reason8.ai, USA) ran at 3.84 m/s. View a video summarizing the competition and showing simulations of the submitted models.

Stanford Health++ Hackathon

The Mobilize Center is excited to be a co-sponsor for the Stanford Health++ Hackathon. This event brings together engineers, designers, healthcare professionals and business experts to collaborate, design and create models and solutions for healthcare affordability here in the US and internationally. This two-day health hackathon will take place October 21-22, 2017 at Stanford University, Huang Engineering Center. Click here to register to participate, mentor, or pitch a need. Featured speakers include Lloyd Minor, Dean of Stanford University School of Medicine.

Highlights of “Data Science in Biomechanics” Thematic Session at ISB 2017

The Mobilize Center organized the thematic session on “Data Science in Biomechanics” at the 26th Congress of the International Society of Biomechanics (ISB) in July. The thematic session generated a lot of interest with over 160 people in attendance. The talks highlighted various data science topics including: 1) how large data from wearables, clinics, and research labs are leading to new actionable insights, and 2) how data science tools can help generate new knowledge that conventional statistical methods may miss.

  • Abstracts
  • Slides: “The Mobilize Center: Accelerating Movement Science with Big Data” by Eni Halilaj
  • Slides: “Transforming Gait Research: Advancing a Worldwide Research Network and Biomechanical Database” by Sean Osis and Reed Ferber
  • Slides: “Long-term Monitoring of Lower Limb Joint Loads Using Wearable Sensors: Application in Sports and Orthopaedics” by Thor Besier, et al.

First Annual eWear Student Symposium

Mobilize Center postdoc Łukasz Kidziński is co-organizing the first annual eWear student symposium taking place Thursday September 7, 2017 at Stanford University. This one day symposium will highlight research on the impact of wearable technology and inspire discussion, generation of ideas, and opportunities for collaboration. Keynote speakers include: Dr. Bill Liu, Founder and CEO of Royole Corporation; Michael S. Eggleston from Bell Labs; and Hind Hobeika, inventor and developer of the first version of Instabeat. Mobilize Center graduate students Tim Althoff and postdoc Jessilyn Dunn will also be speaking at the symposium. Read more and register.

Analysis of Physical Activity Data from Over 700,000 Worldwide Users Reveals Physical Activity Inequality

Mobilize Center faculty member Jure Leskovec and colleagues published a paper in Nature this week where they analyzed physical activity data from over 700,000 users of a smartphone fitness app. Their analysis shows that the disparity of physical activity distribution within a country, what they refer to as “activity inequality,” is a better predictor of obesity levels within a country than average activity levels. In other words, the greater the activity inequality or activity disparity is in a country, the higher the prevalence of obesity. They also identify associations between a city’s walkability and activity inequality. Visit http://activityinequality.stanford.edu/ to access findings from the study, including data and code. Read more.