[ad_1]
BOULDER, Colo., June 16, 2021 / PRNewswire / – DrivenData, the host of data science competitions that advance solutions for social good, and HeroX, the social network for innovation and the world’s leading platform for participatory solutions, announced today They are the winners of the third and final algorithm sprint of the Differential Privacy Temporal Map Challenge, which was sponsored by the Public Safety Communications Research Division (PSCR) of the National Institute of Standards and Technology (NIST).
DrivenData and HeroX
With a purse totaling $ 161,000 on the whole challenge, the announcement today of the third algorithmic sprint proposed $ 25,000 to the first place winner. The “N – CRiPT” team, a group of researchers in differential confidentiality of the national university of singapore, Alibaba Group, got the first place. Their aim was to integrate differential confidentiality into a practical framework. The winner of the second place was the team “Minutemen”, a group of graduate students on differential protection of the University of Massachusetts Amherst.
The aim of this award was to create synthetic data that preserves the characteristics of a dataset containing temporal and geographic information. Synthetic data has the ability to offer better privacy protection than traditional anonymization techniques. Differently private synthetic data can be shared with researchers, policy makers and even the public without risking exposing individuals in the original data. However, synthetic records are only useful if they preserve the trends and relationships in the original data.
Participants in this challenge were tasked with developing algorithms that anonymize datasets while maintaining a high level of precision. This ensures that the data is both private and useful. The best competitors of the final sprint have demonstrated algorithms that produce records with the two more confidentiality and greater precision than typical sub-sampling techniques used by many government agencies to publish records.
The first sprint featured data captured from 911 calls in Baltimore, Maryland completed in one year. Participants in this sprint were tasked with developing anonymization algorithms designed to generate privatized datasets using the number of incidents reported monthly for each type of incident by neighborhood. The winners have been announced here.
The second sprint used demographics from the US Census Bureau’s American Community Survey which interviewed people in various US states from 2012 to 2018. The data set included 35 different survey characteristics (such as the age, sex, income, education, work and health insurance data.) for each individual interrogates. Simulated longitudinal data was created by linking different individual records over several years, which made it more difficult to protect the privacy of each simulated person. To succeed in this sprint, participants had to create de-identification algorithms by generating a set of synthetic and privatized survey records that most accurately preserved the patterns of the original data. The winners have been announced here.
The third sprint focused on taxi journeys taken in Chicago, Illinois. Because the sprint focused on protecting taxi drivers rather than their journeys, competitors had to ensure confidentiality of up to 200 records per individual driver, a very difficult issue. They have been assessed in 77 Chicago community areas. The anonymized synthetic data was needed to preserve the characteristics of taxi trips in each community area, patterns of movement between communities, as well as the characteristics of the population of the taxi drivers themselves (typical times and places of work). The first two winning teams were each able to produce synthetic data that offered very strong privacy protection and were also Following accurate for analysis than data protected by traditional privacy techniques such as downsampling.
Challenge participants are now eligible to win up to $ 5,000 to create and execute a development plan that further improves the code quality of solutions and advances their usefulness to the public safety community. Participants can also win the Open Source Prize, a $ 4000, by publishing their solutions in an open source repository. The winning solutions will be those that respect differential privacy after being uploaded to an open source repository.
About DRIVENDATA
DrivenData is a social enterprise dedicated to bringing the data tools and methods that transform the industry to the world’s greatest challenges. As part of this work, DrivenData’s competition platform channels the skills and passion of data scientists, researchers and other quantitative experts to create solutions for social good. These online machine learning challenges are designed to engage a large community of experts, connect them to real data problems, and highlight their best solutions.
About HEROX
HeroX is a social network of participatory innovation and human ingenuity, co-founded in 2013 by serial entrepreneur, Christian Cotichini and the founder and futurist of XPRIZE, Pierre Diamandis. HeroX offers an easy-to-use, turnkey platform that supports anyone, anywhere to solve everyday business and global challenges using the power of the crowd. Uniquely positioned as a social network for innovation, HeroX is the only place you can create, grow, and organize your own audience.
Media contact:
Alexandra Pony
[email protected]
250.858.0656
To learn more about the eligibility requirements, visit challenge.gov, and for more information about the challenge, visit DrivenData.org.
NIST, a non-regulatory agency of the U.S. Department of Commerce, promotes U.S. industrial innovation and competitiveness by advancing measurement science, standards, and technology in ways that enhance economic security and improve our quality. of life. To learn more about NIST, visit NIST.gov.
Media contacts:
To arrange an interview and / or any media inquiries with NIST, please contact Jennifer huergo at (202) 309-1027 and [email protected].
SOURCE DrivenData; HeroX
[ad_2]
