University of Washington AI Project Takes Madrona Prize At Industrial Affiliates Day

Photo: CoAI team, Joseph Janizek and Gabriel Erion with Tim Porter at the Allen School

Madrona awarded the 14th Annual Madrona Prize to the CoAI team at the University of Washington’s Paul G. Allen School of Computer Science & Engineering. CoAI: Cost Aware Artificial Intelligence for Health Care applies ML to help healthcare professionals use accurate predictive models in time-sensitive and potentially life-threatening situations. The field of cost-sensitive ML builds algorithms that automate the feature selection step, automatically choosing the best subset of input variables to make a high-accuracy prediction. CoAI applies this field to the clinical setting — where “cost” is time — and enables, for instance, an EMT to run the appropriate predictive model while in the ambulance ride to the ER, rather than losing critical minutes after the patient arrives.

The team of consisted of PhD/MD graduate students Gabriel Erion and Joseph Janizek with MDs, Carley Hudelson and Nathan White and the head of UW’s Laboratory for Artificial Intelligence for Medicine and Science, Professor Su-In Lee.

The Madrona Prize is awarded at the end of the Allen School’s annual Industry Affiliates Research Day. The event includes technical talks throughout the day and culminates in an open house and poster session that showcases the latest research projects and papers being pursued by faculty and students at the school. The Madrona Prize has been awarded for 13 straight years and goes to the project that combines excellent research with what we feel is the greatest commercial potential. Since Madrona’s inception more than two decades ago, Madrona has funded 18 companies out of the Allen School. These companies include Impinj (NAS:PI), SkyTap and Turi (acquired by Apple), and most recently, OctoML, a company based on the TVM research project that won a Madrona Prize in 2017.

“The Allen School at the UW is an incredibly important resource for our region and as the school has grown and actively attracted researchers from many different areas, we have seen the breadth and depth of innovation grow,” said Tim Porter managing director, Madrona Venture Group. “Talking with the students during the research day is truly one of the highlights of our year, and we are both excited and inspired by all of the innovative projects we saw.”

Each year, the Madrona committee also awards runner-up prizes. This year the runners up were:

Runners Up
AuraRing: Precise Electromagnetic Finger Tracking via Smart Ring
Farshid Salemi Parizi, Eric Whitmire, Alvin Cao, Tianke Li, Ishan Chatterjee
Advisor: Shwetak Patel

HomeSound: Exploring Sound Awareness In The Home For People Who Are Deaf And Hard Of Hearing
Dhruv Jain, Kelly Mack, Steven Goodman
Advisors: Leah Findlater and Jon Froehlich

Molecular tagging with nanopore-orthogonal DNA strands
Katie Doroschak, Karen Zhang, Melissa Queen, Aishwarya Mandyam, Jeff Nivala
Advisors: Karin Strauss and Luis Ceze

For past winners visit click here.

And the Madrona Prize goes to . . .

(pictured Madrona team, Hank Levy, Runners Up and Madrona Prize Winners)

Last night, Madrona awarded the 12th Annual Madrona Prize to the team at the University of Washington Paul G. Allen School of Computer Science and Electrical Engineering that is building the LabintheWild platform. Working with assistant professor Katharina Reinecke, the team is enabling scientists from all over the world to study how culture influences people’s preferences and abilities while interacting with technology. The LabintheWild platform uses human curiosity to build a thriving community that is helping to increase the understanding of cultural differences. Additionally, the prize was shared with Augury, a program that leverages the insight from LabintheWild to predict how a website will be perceived in different cultures and countries.

(pictured Eunice Jun and Manuel Nordhoff)

This year the Madrona Prize went to the team that is working on LabintheWild, which has been used by nearly 4 million participants and is a platform for scientists to test behavioral response and perceptions across the globe. Users take these surveys and immediately see how their response compares with other people in their region and beyond. Augury shared this prize and has leveraged learnings from the LabintheWild platform to create a system that enables website builders to test the response to their design by country. The team on this project who will share the prize are Eunice Jun, Nigini Oliveira, and Manuel Nordhoff. The prize is a cash award that goes directly to the team involved in directing and conducting the research.

The newly named Paul G. Allen School has a long history of research that improves lives and delivers business and research growth to the region

Tim Porter, Managing Director

The Runner Up projects were focused in the areas of machine learning and VR. The open source TVM framework simplifies and optimizes running deep learning models on a variety of hardware. A Unified Approach to Interpreting Model Predictions tackles explaining the output from ML models, using other computational methods. Understanding the “why” behind the insights produced by AI and ML is becoming increasingly important across businesses and research institutions. And, Real-Time VR Video Processing with the Hardware-Friendly Bilateral Solver optimizes the processing of VR video in the cloud, a commercial problem we are well acquainted working with VR companies such as Pixvana.

“The newly named Paul G. Allen School has a long history of research that improves lives and delivers business and research growth to the region,” said Tim Porter managing director, Madrona Venture Group. “The UW is an invaluable resource to our technology ecosystem and we are looking forward to the increased impact that the new Bill and Melinda Gates Center will make possible. Research such as LabintheWild and the runners up are key components to what makes the Seattle technology industry so vibrant.”

“The Allen School is lucky to be part of a region teeming with global technology companies who partner with us, and our yearly research day is an opportunity to showcase our latest research innovations and our talented students for this community,” said Hank Levy, Director of the Allen School. “We really appreciate Madrona’s partnership over the years and their role in our ecosystem. They encourage our students through the Madrona Prize and they amplify our impact by starting and building companies based on technologies we develop.”

The Madrona Prize is awarded at the end of the Allen School Industrial Affiliates day. The Industrial Affiliates brings together all the top research projects and papers being pursued by professors, graduate and undergraduate students at the school. The Madrona prize goes to the research project with the most commercial potential. Since Madrona’s inception, more than two decades ago, Madrona has funded 17 companies out of the University of Washington. These companies include Impinj (NAS:PI), SkyTap and Turi (acquired by Apple.)

Each year, the Madrona committee also awards Runner up prizes. This year the Runners up were:

Runners Up

A Unified Approach to Interpreting Model Predictions

Scott Lundberg

Adviser: Su-In Lee

TVM: End-to-End IR Stack for Deep Learning Systems

Tianqi Chen; Thierry Moreau; Haichen Shen

Advisers: Luis Ceze, Carlos Guestrin, Arvind Krishnamurthy

Real-Time VR Video Processing with the Hardware-Friendly Bilateral Solver

Amrita Mazudar, Armin Alaghi

Advisers: Luis Ceze, Mark Oskin