News Release

KDD teams tap advanced data science to tackle societal challenges

Prizes offered for innovative solutions to problems including transportation, malaria

Business Announcement

Association for Computing Machinery

ACM KDD 2019

image: ACM KDD 2019 will take place in Anchorage, Alaska -- August 4-8, 2019. view more 

Credit: KDD

KDD 2019, the premier interdisciplinary data science conference, announced KDD Cup 2019, the 23rd annual data mining and knowledge discovery competition organized by the ACM Special Interest Group on Knowledge Discovery and Data Mining. For the first time, this year's competition will feature three distinct competition tracks, each presenting a real-world challenge that participants look to solve using machine learning and artificial intelligence. Winners will be announced at KDD 2019, held Aug. 4-8, 2019 in Anchorage, Alaska.

"For more than two decades, the KDD Cup has seen the brightest minds in machine learning tackle some of the most difficult and unique challenges in the world," said KDD Cup Chair Taposh Dutta-Roy. "We're thrilled that this year's competition features three separate tracks, challenging a wide range of diverse thinkers and leaders in the field of machine learning, and enhancing visibility for the powerful impact of data science in an array of sectors."

The three competition tracks include:

  • "Research for Humanity" Reinforcement Learning Competition Track (Humanity RL Track)--The first-of-its-kind Humanity RL Track, sponsored by IBM Research Africa and Hexagon-ML, requires participants to apply machine learning tools to determine novel solutions that could impact malaria policy in sub-Saharan Africa. Specifically, the competition looks at how combinations of interventions that control the transmission, prevalence and health outcomes of malaria infection should be distributed in a simulated human population. First place receives $5,000.
  • Regular Machine Learning Competition Track (Regular ML Track)--Sponsored by Baidu, this year's regular ML track has participants submit a context-aware multi-modal transportation recommendation in the form of a travel plan. Submitted travel plans will consider various unimodal forms of transport, like walking, driving and public transit, connecting them under various contexts to optimize transportation recommendations across a variety of users and spatiotemporal contexts. First place receives $10,000.
  • Automated Machine Learning Competition Track (Auto-ML Track)--In the Auto-ML challenge, sponsored by 4Paradigm, ChaLearn and Microsoft, participants are invited to deploy Auto-ML solutions to binary classification problems for temporal relational data. Each team is given five public datasets to develop Auto-ML solutions. These solutions will be evaluated with five unseen datasets without human intervention, and the winners will be chosen based on the final rankings of the datasets. First place receives $15,000.

More than 3,000 teams participated in this year's competition, with winners of each track selected entirely by a non-human, automated process. The previous 22 KDD Cup competitions have covered a wide-range of challenging predictive problems provided by industrial, research, educational and other organizations. Among those problems are forecasting air quality indices, estimate highway tollgates traffic flow, predicting course drop-outs for college students, and predicting followers and click-through rates to improve user engagement with social network platform content. Last year's KDD Cup winning team hailed from Beijing University of Posts and Telecommunications and Central South University in Changsha, China.

"We have carefully reviewed past KDD Cup challenges and summarized takeaways from those competitions," said KDD Cup Chair Wenjun Zhou, University of Tennessee. "KDD Cup has gained popularity over the years, and we are always looking for ways to challenge the next round of participants. This year, we aimed to introduce novelty to the competition and called for proposals that would reflect the most interesting developments in data science. With more than 5,000 submissions this year, we are excited to unveil this year's outstanding projects."

KDD 2019 brings together leading experts in the world of data science and artificial intelligence to share their latest research results and apply recent findings to the challenges facing an array of industries. The event is comprised of workshops, tutorials and designated special theme days, which highlight machine learning applications for environmental sustainability, healthcare and deep learning.

"This year's competition includes different tracks that offer a platform for teams and individuals to share, collaborate and improve the overall knowledge base in machine learning and artificial intelligence," said KDD Cup Chair Iryna Skrypnyk, Pfizer. "We're pleased to unveil a brand new 'Research for Humanity' track that deals with a real-world crisis in Africa and a new Auto-ML track, while also continuing to offer the popular Regular ML track that KDD Cup has prided itself on for more than 20 years."

KDD 2019 will be held at the Dena'ina Convention Center and William Egan Convention Center. For more information, visit: https://www.kdd.org/kdd2019/.

###

About ACM SIGKDD:

ACM is the premier global professional organization for researchers and professionals dedicated to the advancement of the science and practice of knowledge discovery and data mining. SIGKDD is ACM's Special Interest Group on Knowledge Discovery and Data Mining. The annual KDD International Conference on Knowledge Discovery and Data Mining is the premier interdisciplinary conference for data mining, data science and analytics.

For more information on KDD, please visit: https://www.kdd.org/.

Follow KDD on:

Facebook -- https://www.facebook.com/SIGKDD
Twitter -- https://twitter.com/kdd_news
LinkedIn -- https://www.linkedin.com/groups/160888/

Media Contact:

Havas Formula for KDD 2019
KDD@havasformula.com
619-234-0345


Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.