All posts by admin

OpenSim Webinar – Inverse Kinematics: A Bayesian Versus Least-Squares Approach

Learn the basics of Bayesian inference and how it can be used for inverse kinematics.

DETAILS

Title: Inverse Kinematics: A Bayesian Versus Least-Squares Approach
Speaker: Todd Pataky, Kyoto University
Time: Monday, July 20, 2020 4:00 p.m. Pacific Time
Registration: Registration is free but required, as space is limited. Click here to register.

ABSTRACT

When estimating joint kinematics from a set of noisy marker measurements, most inverse kinematics (IK) approaches aim to minimize errors by distributing them amongst the markers in a least-squares sense. This talk will describe how Bayesian calculations can be used to maximize the probability that a specific set of joint angles would produce the observed marker positions. This computation of marker positions from joint angles is referred to as a forward kinematics model or forward model. Preliminary Bayesian IK results show order-of-magnitude improvement over least-squares estimates, with Bayesian IK outperforming least-squares IK in over 90% of a large number of random simulations involving planar rotations, with an approximately tenfold decrease in rotation estimate error. This talk will assume no knowledge of Bayesian statistics, and will work from fundamental Bayesian concepts to simple IK models, then finally to explain why Bayesian IK appears to outperform least-squares approaches.

Applications now open for the mHealth Training Institute 2.0.

Applications are now being accepted for the 2020 mHealth Training Institute (mHTI) to be held at the University of California, Los Angeles from July 26 – 31, 2020. The week-long, residential program blends “deep dives” into the latest mHealth technologies and methodologies with team-science projects to provide the selected scholars with the transdisciplinary knowledge, competencies and collaborative networks required to tackle “wicked” healthcare problems.  The mHTI is funded by a NIH/OBSSR/ODP/NIDA training grant (R25DA038167) and administratively supported by the NIH Center for Excellence on Mobile Sensor Data-to-Knowledge (MD2K).

Version 2.0 builds on the experiences of previous mHTIs and incorporates feedback and suggestions from past participants.  Scholars from earlier institutes have highlighted a wide range of benefits, describing the mHTI as “transformative” and “a seminal experience in my professional development” with a “wonderful atmosphere and congeniality among scholars” and “terrific and generous faculty.” The mHTI derives from the mentoring provided by a core group of experienced faculty; mHealth thought-leaders who are dedicated to developing the next generation of mHealth scientists. Details of previous mHealth Training Institutes including scholars, faculty, and programs are available at https://mhti.md2k.org/.

Apply now!

Key Dates

Application Deadline11:59 PM, EASTERN TIME, Sunday, March 1, 2020.

Notifications of acceptanceApril 15, 2020.

Questions:  info@MD2K.org

Tim Althoff receives the 2019 SIGKDD Dissertation Award

We are delighted to announce that our former graduate student, Tim Althoff, received the 2019 SIGKDD (Conference on Knowledge Discovery and Data Mining) Dissertation Award. Dr. Althoff was recognized for his work on data mining to improve health and well-being. While Dr. Althoff was at Stanford University, he was advised by Mobilize Center faculty member Jure Leskovec. Now Dr. Althoff is an assistant professor at the University of Washington. You can find the slides from his talk at SIGKDD here. All the code and data for the presented papers are also available:

Podcast Features Women Leaders at Intersection of Data Science and Genetics

Listen to women leaders across the data science profession, as they share their advice, career highlights, and lessons learned along the way. In season one, the Women in Data Science (WiDs) podcast takes a unique look into the lives of two women working at the intersection of data science and genetics:

  • Chiara Sabatti, professor of biomedical data science and of statistics at Stanford University, discusses trends in data science in genetics. Watch now
  •  Nilah Monnier Loannidis, a postdoc in the Department of Biomedical Data Science at Stanford University, discusses the role of data in her career and new ways to collect data. Watch now

This podcast is brought to you by the Stanford Institute for Computational & Mathematical Engineering (ICME) and the Stanford School of Engineering. Support for this podcast and other Women in Data Science initiatives has been provided by Intuit, Microsoft, SAP, Walmart Labs, and Western Digital. The Mobilize Center is one of the initial sponsors of the Women in Data Science Conference. To listen to all the WiDS episodes, click here.

“Our Voice” citizen-science project featured in 3M Particles website

Mobilize Center faculty member Dr. Abby King and her team’s “Our Voice” citizen-science project is featured in 3M’s Particles website. The project engages community members in improving their neighborhoods and making them more conducive for physical activity. Using a tablet-based app, people can take geo-coded photos and videos to highlight for policy makers and community leaders which neighborhood areas are in greatest need of improvement. This project is synergistic with Dr. King’s earlier study published in Nature with other Mobilize Center collaborators that analyzed data from 720,000 people and found a correlation between a location’s walkability and activity levels. Read more

The Mobilize Center Funds 6 Students to Attend This Year’s Machine Learning for Healthcare Conference

This year the Mobilize Center supported 6 students to attend The Machine Learning for Healthcare Conference (MLHC) at Stanford University on August 17-18, 2018.

Congratulations to Jeeheh Oh, Jacob Fauber, Bryan Lim, Xinyuan Zhang, Bryce Woodworth, and Zelun Luo! We welcome you to read their papers below:

Bryan Lim, Oxford University. Disease-Atlas: Navigating Disease Trajectories using Deep Learning

Bryce Woodworth, UC San Diego. Preference Learning in Assistive Robotics: Observational Repeated Inverse Reinforcement Learning

Jacob Fauber, UC Riverside. Modeling “Presentness” of Electronic Health Record Data to Improve Patent State Estimation

Jeeheh Oh, University of Michigan. Learning to Exploit Invariances in Clinical Time-Series Data using Sequence Transformer Networks 

Xinyuan Zhang, Duke University. Multi-Label Learning from Medical Plain Text with Convolutional Residual Models

Zelun Luo, Stanford University (PI: Fei Fei). Computer Vision-based Descriptive Analytics of Seniors’ Daily Activities for Long-term Health Monitoring