San Francisco, CA

October 15-17, 2017

101 Howard Street | view map >

Drawing leaders of industry and academia to explore the latest theoretical advances and technological applications in data science in order to promote the next generation of cross-disciplinary research.

Get notified about conference updates


Some of Our Speakers

What to Expect

Limited to 350 participants, conference attendees can expect to learn from an invited group of cutting edge researchers from academia and industry who address the latest advancements across the spectrum of data science. This intensive three day conference will include workshops, poster presentations, parallel sessions and plenary talks. In addition to the traditional open exchange of ideas of an academic conference, a key outcome of this conference is the forging of new collaborations across the traditional divide between industry and academia.

Tracks may include

Workshops will include

Ticket Info

Academic Registration:

Academics include full-time faculty members and researchers, postdoctoral students or full-time graduate students.

Association Discount (ACM, ASA, SIAM)
Registration includes: Access to all sessions, complimentary breakfasts, coffee, plenary talks, welcome reception, poster session + reception
$50 Additional for access to Sunday workshops
Academic Pass
Registration includes: Access to all sessions, complimentary breakfasts, coffee, plenary talks, welcome reception, poster session + reception
$50 Additional for access to Sunday workshops

Industry Registration:

Industry Pass
Registration includes: Access to all sessions, complimentary breakfasts, coffee, plenary talks, welcome reception, poster session + reception
$100 Additional for access to Sunday workshops
One-day pass
Registration includes: Access to one day of concurrent sessions, complimentary breakfast and coffee, and one plenary session


Sunday, October 15

12:30 - 1:00 PM

☕ Coffee, Refreshments, and Registration

1:00 - 3:00 PM

Workshop Session I

Introduction to Deep Learning with Python I
Jeremy Howard, fast.ai

Designed Experiments for Data Scientists
Nathaniel Stevens, USF

3:30 - 5:30 PM

Workshop Session II

Introduction to Deep Learning with Python II
Yannet Interian, USF

Network Analytics
James D. Wilson, USF

5:30 - 6:00 PM

🍷🍺 Beer, wine and Refreshments

6:15 - 7:15 PM
Keynote Talk - Large-scale Machine Learning: Deep, Distributed and Multi-Dimensional
Animashree Anandkumar, Amazon AWS
7:15 - 10:30 PM
Offsite: Hacknight
Michael Terry, Lawfty
Eat, drink and "hack the NFL" with fellow coders and data scientists in an open coding environment. Prizes will be awarded to the best and most unique predictive models - Hosted by LAWFTY
Club Quarters Hotel (opposite Embarcadero Center)
424 Clay Street
San Francisco, CA 94111

Monday, October 16

8:30 - 9:00 PM

☕ Coffee, Refreshments, and Registration

9:00 - 10:30 AM

Concurrent Sessions

Networks I
Chair: Peter Mucha, UNC Chapel Hill

Layer-Coupled centralities and embeddings for multilayer and time-varying networks
Dane Taylor, U. at Buffalo

Counter adversarial community detection
Ali Pinar, Sandia National Labs

Dynamic networks of animal behavior
Tanya Berger-Wolf, U. Illinois at Chicago

Sifting through measures on networks: from a theoretical framework to an empirical guide
Tina Eliassi-Rad, Northeastern

Experimental Design I
Chair: Nathaniel Stevens, USF

Practical testing
Nick Ross, USF

Firefox observational study: Do users become more engaged as a result of using add-ons?
Ben Miroglio, Mozilla

Thompson sampling for infinite armed bandits
Steven Scott, Google

The power of A/B testing under network interference
James D. Wilson, USF

Compressed Sensing and Machine Learning I
Chair: Deanna Needell, UCLA

Debiasing 1-step matrix completion
Yaniv Plan, UBC

Deep Learning Meets Matrix Factorization
Blake Hunter, Claremont McKenna College

Personalization outside the box
Sonya Berg, Stitchfix

Relative error tensor low rank approximation
David Woodruff, IBM Almaden

10:30 - 10:45 AM

☕ Coffee and Refreshments

10:45 - 11:15 AM

Opening remarks

11:15 AM - 12:15 PM

Keynote Talk – Machine Learning: Perspectives and Challenges

Michael I Jordan, UC Berkeley
12:15 - 1:30 PM

🍔 Lunch Break

Mentoring lunch (invited guests)

1:30 - 3:00 PM

Concurrent Sessions

Networks II
Chair: Peter Mucha, UNC Chapel Hill

Avoiding the resolution parameter in community detection
John Palowitch, Google

Self-diagnosing networks
Nina Fefferman, U. Tennessee

Higher-order clustering coefficients
Austin Benson, Cornell

Parameter inference and model selection for mechanistic network model
JP Onnela, Harvard

Data Science for Social Good I
Chair: Rachel Thomas, fast.ai

Bridging the data gap between Silicon Valley and the rest of the world: Using small data to do good
Roberto Sanchis Ojeda, Netflix

Data science for a skeptical public: creating a field experiment to inform police stop data collection
Eric Giannella, CA DOJ

From Theory to Practice: A case study in putting deep learning models in the hands of radiologists
Peter Bull, DrivenData

Bias in machine learning
Lisa Green, Domino Labs

Compressed Sensing and Machine Learning II
Chair: Mary Wootters, Stanford

Ben Recht, UC Berkeley

If it ain’t broke, don’t fix it: Sparse metric repair
Anna Gilbert, U Michigan

Compressed sensing using generative model
Eric Price, UT Austin

Computing a quantity of interest from observational data
Simon Foucart, Texas A&M

Machine Learning and Health I
Chair: Gari Clifford, Emory University and Georgia Institute of Technology

Iterative random forests to discover predictive and stable high-order interaction
Karl Kumbier, UC Berkeley

The cGERGM - A generative model for correlation networks of the brain
Paul Stillman, OSU

Overcoming the embodiment problem of application of AI to medicine via streaming predictive analytics and clinician-in-the-loop control
Shamim Nemati, Emory University

Machine Learning for Precision Medicine: Challenges and Promises
Stathis Gennatas, University of California, San Francisco

3:00 - 3:30 PM

☕ Coffee and Refreshments

3:30 - 5:00 PM

Concurrent Sessions

Deep Learning I
Chair: Jeremy Howard, fast.ai

Deep learning for e-commerce personalization
Wei Dai, Reflektion

Deep learning for classifying advertisement content in online advertising
Daniel Austin, AppNexus

Backing off towards simplicity - why complex isn't necessarily better in deep learning
Stephen Merity, Salesforce

Learning geo-temporal patterns in location data: Application to preventing account takeovers
Lenny Evans, Uber

Search and Optimization I
Chair: Sriram Sankar, Facebook

Attribution of offline activity to online ad spending
Michael Terry, Lawfty

Search ranking systems at Pinterest
Zhongxian Chen, Pinterest

Carl Steinbach, LinkedIn

Recommendation systems to intelligent bots - a new age of conversational interfaces
Mitul Tiwari, Ruperbot

Machine Learning Applications I
Chair: Matthew Herman, InView

Making social-network data sets more human: a topological approach
Cynthia Phillips, Sandia Labs

Accounting for linguistic diversity
Rachael Tatman, University of Washington

Deep Learning at scribd
Kevin Perko, scribd

Voting data and the use of geospatial analytics
Ben Uminsky, LA County Registrar

5:00 - 6:30 PM

🍷🍺 Poster Session, Beer, and Wine

7:00 - 9:00 PM

Offsite I: Dinner for Conference Organizers and Keynote Speakers
Offsite II: Offsite Happy Hour for conference guests

Tuesday, October 17

8:00 - 9:00 AM

☕ Coffee, Refreshments, and Registration

9:00 - 10:30 AM

Concurrent Sessions

Networks III
Chair: Peter Mucha, UNC Chapel Hill

Understanding individual variability through data-driven modeling of human brain dynamics
Sarah Muldoon, U. at Buffalo

Information flow and prediction limits in online social networks
James Bagrow, U. Vermont

Multi-layer planted cluster detection via intersection and convex aggregation methods
Rajmonda Caceres, MIT Lincoln Labs

Exploiting user relationships to accurately predict preferences in large-scale social networks
Jennifer Neville, Purdue

Machine Learning and Health II
Chair: Gari Clifford, Emory University and Georgia Institute of Technology

Using machine learning to improve case definition development in a Canadian primary care electronic medical record database
Tyler Williamson, U Calgary

Machine learning for FDA approved consumer level point of care diagnostics - the wisdom of algorithm crowds
Gari Clifford, Emory U

Accurate and interpretable machine learning algorithms
Gilmer Valdes, UCSF

Nonlinear EEG Analysis for Early Detection of Autism Spectrum Disorder: A data-driven approach"
William Bosl, University of San Francisco

Sparse Data Recovery and Machine Learning
Chair: Xuemei Chen, USF

Distributed estimation using the median of means-estimator
Nate Strawn, Georgetown

Fast high-resolution EEG source imaging
Jing Qin, Montana State

Data adaptive kernel maximum mean discrepancy
Alexander Cloniger, UC San Diego

Multiscale adaptive approximations to data and functions near low-dimensional sets
Wenjing Liao, Georgia Tech

10:30 - 11:00 AM

☕ Coffee and Refreshments

11:00 AM - 12:00 PM

Keynote Talk – Smartphone-based Digital Phenotyping

JP Onnela, Harvard
12:15 - 1:30 PM

🍔 Lunch Break

1:30 - 3:00 PM

Concurrent Sessions

Experimental Design II
Chair: Nathaniel Stevens, USF

Online experimentation: the promise, perils, and pitfalls
Roger Longbotham, Microsoft

Designing optimal incentives for a ridesharing marketplace
Jose Abelenda, Lyft

Non-confounding alternatives to regular 2^{k-p} fractional factorial designs
Douglas Montgomery, ASU

A meta-analysis of response surface studies
David Edwards, VCU

Deep Learning II
Chair: Jeremy Howard, fast.ai

Deep learning at Instacart
Jeremy Stanley, Instacart

Graph mining and deep learning for e-commerce personalized search and recommendation
Weizhi Li, Reflektion

Danqi Chen, Stanford

Doing strange things with attention
Colin Raffel, Google Brain Resident

Machine Learning Applications II
Chair: Matthew Herman, InView

Using Kafka and Spark to explore the simultaneous localization and mapping (SLAM) problem for robotics for autonomous vehicles
Jay White Bear, MIT Lincoln Labs

On using a 64x64 sensor to optically compute the first layers of a CNN
Matthew Herman, InView

Compression networks with super nodes
Natalie Stanley, UNC Chapel Hill

3:00 - 3:30 PM

☕ Coffee and Refreshments

3:30 - 5:00 PM

Concurrent Sessions

Networks IV
Chair: Peter Mucha, UNC Chapel Hill

Communities in multilayer networks: CHAMP
Peter Mucha, UNC Chapel Hill

Searching large networks using multi-resolution neighborhood structure
Blair Sullivan, NCSU

Near-optimal and practical algorithms for graph scan statistics
Anil Vullikanti, Virginia Tech

Data Science for Social Good II
Chair: Rachel Thomas, fast.ai

Using deep learning and Google street view to estimate the demographic makeup of the US
Timnit Gebru, Microsoft Research

Ethics for powerful algorithms
Abe Gong, Founder and CEO of Superconductive Health

Transparency in machine learning: understanding feature importance in image classification models
Sara Hooker, Brain Resident at Google

On learning AI: The myth of innate ability in tech
Omoju Miller, GitHub

Search and Optimization II
Chair: Sriram Sankar, Facebook

Monte carlo simulation and business risk
Jacob Pollard, Lawfty

Measuring value: be careful what you optimize for
Bonnie Barrilleaus, LinkedIn

Pricing Challenges at Airbnb
Li Zhang, Airbnb

Search as a service @ scale
Sriram Sankar, Facebook

5:15 - 6:30 PM
Henry Humadi, VungleChristina Choi, Eventbrite
Kimberly ShenkClaire Lebarz, Airbnb
Jeremy Stanley, InstacartMike Brzustowicz, Sandia National Labs

🍷🍺 Panel: The State of Data Science in Industry

Moderator: Jeremy Howard, fast.ai


AWS     CapitalOne Vungle ValiMail
   Reputation.com Cogitativo

In cooperation with


Interested in being a sponsor? Please see our sponsorship page for more information

Hotel Information

Courtyard by Marriott San Francisco Downtown
299 2nd St, San Francisco, CA 94105
(0.5 miles away)
The Handlery
351 Geary Street, San Francisco, CA 94102
(0.8 miles away)

Organizing Committee