At the Stanford Women in Data Science Conference (WiDS) today, the winners of the 2020 WiDS Datathon were announced. The datathon, which focused on patient health through data from MIT’s GOSSIS (Global Open Source Severity of Illness Score) consortium resulted in three winning teams: 1st place: Women Power, from Israel; 2nd place: nullset, from Ukraine; 3rd place: Prevision.io from France. These teams and their fellow competitors used various tools to surface patterns and insights from Intensive Care Unit (ICU) data, developing models to predict survival. These winners were among 951 teams from approximately 85 countries that competed in this year’s WiDS datathon.
One of the winning teams has been competing in the WiDS datathon since its inception in 2018. “We first teamed up to compete three years ago and actually placed 5th at the WiDS Datathon in 2018,” said Iryna Ivanenko from team nullset. “That was a good start, but today we’re so happy to finish in second place out of a field of such impressive competition. As a Data Science Lead at Women Who Code Kyiv, which organized the WiDS Kyiv 2018 event, I’ve seen first hand how these WiDS events encourage women in the field.”
The WiDS datathon encourages women to hone their data science skills through a predictive analytics challenge focused on social impact, bringing people together across borders to work in teams, solving global challenges. Participants received access to data from more than 130,000 hospital ICU visits, and were able to compare the performance of their models through a public leaderboard on Kaggle. With approximately 500,000 ICU deaths annually in the U.S. alone, this data exploration has the potential to make significant improvements on a global scale.
“The WiDS datathon has been successful in training and engaging many more women in predictive analytics challenges,” said Karen Matthys, Executive Director, External Partners at Stanford’s Institute for Computational & Mathematical Engineering (ICME) and Co-Director of the WiDS initiative. “WiDS has done a few things differently – including focusing the datathon on a topic with significant social benefit, offering online tutorials and in-person datathon starter workshops in over 20 locations, and by requiring teams to have at least 50 percent women. In 2020, over 80 percent of participants in the WiDS Datathon identify as female–compared to the normal Kaggle competition that has less than 20 percent women. We’re thrilled to see such great diversity!”
This year’s WiDS datathon competitors can also participate in a new second phase to the WiDS datathon, competing for the Excellence in Research Award, which is designed to encourage deeper exploration and the development of collaborative data science innovations. Participants can share more details about their work on the datathon by submitting a one-page paper on their findings by March 31, 2020 here.
The WiDS datathon is part of the WiDS global initiative, which reaches more than 100,000 people worldwide each year through a technical conference at Stanford and at 150+ locations around the world and online, as well as an on-going podcast series. WiDS is spearheaded by the Stanford Institute for Computational and Mathematical Engineering (ICME) and aims to inspire and educate data scientists worldwide, regardless of gender, and to support women in the field.
The WiDS datathon is a collaboration led by the global WiDS team at Stanford ICME, the West Big Data Innovation Hub, and the WiDS datathon committee, and is made possible by the Global Visionary Sponsors: Facebook, Intuit, Walmart Labs, and Wells Fargo. The WiDS Datathon Excellence in Research Award is supported by the WiDS datathon committee and the National Science Foundation Big Data Innovation Hubs.
The Women in Data Science (WiDS) initiative aims to inspire and educate data scientists worldwide, regardless of gender, and to support women in the field. WiDS started as a conference at Stanford in November 2015. Now, WiDS includes a global conference, with approximately 150+ regional events worldwide; a datathon, encouraging participants to hone their skills using a social impact challenge; a podcast, featuring leaders in the field talking about their work, their journeys, and lessons learned; and ongoing education initiatives.