both synchronous and asynchronous parallelization
have been established. However, there are also many
open issues in this field, such as how to select an
appropriate infrastructure and parallelization mechanism given the application and system configuration,
why many papers report linear speedups, but when
the accuracy on real-world workloads, the practical
speed-up is far smaller, why parallelization mechanisms with similar convergence rates perform so differently in practice, and how one conducts proper
comparison and evaluation for distributed machine
learning (for example, benchmark, criteria, system
configurations, and baselines).
Without answers to these important questions,
people can hardly be confident in wide adoption of
distributed machine learning in real applications. The
hope for this workshop was to provide the commu-
nity with deep insights and to substantially push the
frontier of distributed machine learning.
Invited talks were delivered by Alex Smola (Ama-
zon), Joseph E. Gonzales (University of California,
Berkeley), Xiangrui Meng (Databricks), and Christo-
pher Ré (Stanford University). The Distributed Ma-
chine Learning workshop was organized by Tie-Yan
Liu, James Kwok, and Chih-Jen Lin. This summary of
the workshop was reproduced from the technical
report. No report was submitted by the organizers.
The papers presented at the workshop were published
as AAAI Technical Report WS-17-08 in the AAAI Dig-
ital Library and included in The Workshops of the Thir-
ty-First AAAI Conference on Artificial Intelligence: Tech-
nical Reports WS-17-01 – WS- 17-15.
Population health intelligence includes a set of activ-
ities to extract, capture, and analyze multidimen-
sional socioeconomic, behavioral, environmental,
and health data to support decision making to
improve the health of different populations.
Advances in artificial intelligence tools and tech-
niques and Internet technologies are dramatically
changing the ways that scientists collect data and
how people interact with each other and with their
environment. Moreover, the Internet is increasingly
used to collect, analyze, and monitor health-related
reports and activities and to facilitate health-promo-
tion programs and preventive interventions. In addi-
ton, to tackle and overcome several issues in person-
alized health care, information technology will need
to evolve to improve communication, collaboration,
and teamwork among patients, their families, health-
care communities, and care teams involving practi-
tioners from different fields and specialties.
This Health Intelligence workshop follows the success of the earlier workshops held in conjunction
with the 27th, 28th, 29th, and 30th AAAI Conferences on Artificial Intelligence. This joint workshop
brought together a wide range of particpants (about
50 registrants) from the multidiciplinary field of medical and health informatics. Participants were interested in the theory and practice of computational
models of web-based public health intelligence as
well as personalized health-care delivery. The papers
and demonstrations presented at the workshop covered a broad range of disciplines within artificial
intelligence including knowledge representation,
machine learning, natural language processing, pattern recognition, digital imaging, and online social
media analytics. From an application perspective,
presentations addressed topics in epidemiology, environmental and public health informatics, disease surveillance and diagnosis, patient participation, health
behavior monitoring, and disaster management.
The workshop also included four invited talks.
Rumi Chunara (Global Institute of Public Health,
New York University) gave a presentation on the use
of unstructured data in population health. Urmimala
Sarkar, MD (University of California San Francisco
and San Francisco General Hospital) described values
of social media in health IRL applications. Mor Peleg
(University of Haifa) presented her findings from the
MobiGuide Project on how to promote patients’
engangement in their health-care decision-making
process. John H. Holmes (University of Pennsylvania
Hospital) also gave an insightful presentation on AI-driven approaches to data integration.
To promote open debate and exchange of opinion
among participants, the workshop held a panel discussion moderated by David L. Buckeridge and
included Deborah L. McGuinness (Rensselaer Polytechnic Institute), Rumi Chunara (New York University), and José Luis Ambite (University of Southern
California). The major theme of the panel was to discuss the synergy between precision health for individuals and populations.
The Health Intelligence joint workshop was organized by Arash Shaban-Nejad and Martin Michalowski.
This report was written by Arash Shaban-Nejad, Martin Michalowski, David L. Buckeridge, Byron C. Wallace, Michael J. Paul, Szymon Wilk, and John S.
Brownstein. The papers presented at the workshop
were published as AAAI Technical Report WS-17-09 in
the AAAI Digital Library and included in The Workshops of the Thirty-First AAAI Conference on Artificial
Intelligence: Technical Reports WS-17-01 – WS- 17-15
Human-Aware Artificial Intelligence
As AI techniques and systems come into increasing
contact with humans, and into the public conscious-
ness at large, various research issues surrounding such
interactions have come to the fore. Specifically, a key
movement that is underway in the AI community
and the world of technology at large concerns the
notion of humans and machines (AI systems) team-
ing up together to understand data and take deci-