** Location: Room 116 **


08:45     09:00       Opening remarks

Session 1: Modeling models  (Chair: Aparna Lakshmiratan)

09:00     09:20       Invited Talk: You've been using asynchrony wrong your whole life! (Chris Re, Stanford)

09:20     09:40       Contributed Talk: Hemingway: Modeling Distributed Optimization Algorithms

09:40     10:00       Invited Talk: Paleo: A Performance Model for Deep Neural Networks (Ameet Talwalkar, UCLA)

Session 2: Poster Previews  (Chair: Aparna Lakshmiratan)

10:00     10:40       12 x 3 min lightning talks

Posters & Coffee

10:40     11:30       Posters & Coffee

Session 3: Both Ends of the Deep  (Chair: Siddhartha Sen)

11:30     11:50       Invited Talk: Scaling Machine Learning Using TensorFlow (Jeff Dean, Google Brain)

11:50     12:10       Contributed Talk: Demitasse: SPMD Programing Implementation of Deep Neural Network Library for Mobile Devices


12:10     13:30       Lunch 

                                *Optional*: John Langford will give a VW tutorial from 12:30 - 13:20 for those who want to stay/eat in the room! 

Session 4: ML Systems Updates  (Chair: Sarah Bird)

13:30     14:50       Updates from Caffe (Andrew Tulloch), MxNET (Tianqi Chen), Torch (Soumith Chintala), CNTK (Nikos Karampatziakis), VW (John Langford), Decision Service (Siddhartha Sen), and Clipper (Daniel Crankshaw) 

Session 5: End-to-End  (Chair: Li Erran Li)

14:50     15:10       Invited Talk: Optimizing Machine Learning and Deep Learning (John Canny, UC Berkeley & Google Research)

15:10     15:30       Invited Talk: Optimizing Large-Scale Machine Learning Pipelines with KeystoneML (Tomer Kaftan, UW)

Poster & Coffee

15:30     16:30       Posters & Coffee

Session 6: Decisions, decisions... trees  (Chair: Li Erran Li) 

16:30     16:50       Contributed Talk: Yggdrasil: An Optimized System for Training Deep Decision Trees at Scale

16:50     17:10       Contributed Talk: TensorForest: Scalable Random Forests on TensorFlow


17:10     17:15       Closing remarks