GNY, the world’s first decentralized machine learning platform demonstrated its secure data collaboration and machine learning capabilities for analyzing COVID-19 mortality trends in different U.S. cities.
The demo, which used mortality data from the John Hopkins University, showed how different states are able to share their data on a blockchain architecture and use GNY’s machine learning services without risking sensitive data theft.
The data is never stored in a central location, meaning hackers have no server to attack. GNY’s machine learning platform processes the data where it is directly on the chain and shares the results back with the client. This means users worry about neither the data security nor the machine learning algorithms, as GNY offers both.
“This global lockdown has forced us all to think about how our digital infrastructure can support new and better ways to collaborate with data more effectively in a secure environment,” said Cosmas Wong, CEO of GNY. “We are proud to deliver a solution that marries the power of two of the most exciting technologies to date; machine learning and the blockchain.”
In this demo, GNY decentralized a Support Vector Machine (SVM) algorithm on-chain to detect which U.S. cities are behaving irregularly compared to other locations in terms of mortality rates. This information could indicate if certain cities have taken measures that are exceptionally good and should be adopted by others as well, or if some of their measures had catastrophic results and should be reverted immediately.
All of this is done without risking privacy or security. The findings of the ML model is shown in a graph below. For the purposes of this demo, the city names have been anonymized.
The bracketed data on the right represents actual cumulative COVID-19 deaths from May 17th and 18th for 6 US cities. The positive or negative integer above them is the result of the SVM analysis. Positive ones represent cities that are following a pattern that is being observed in most other US cities, in other words progressing in a standard vector, and are represented in red.
The cities showing a negative one value are the outliers or cities whose mortality rates are changing outside the range of their vector are flagged as outliers (blue). 3148 US cities were analyzed using the decentralized SVM model, and 283 were shown to be significant outliers, with a 20%+ variation from the overall vector shown by all cities.
This demonstration previews the full functionality of GNY’s Mainnet, which is launching later this year. The Mainnet will provide businesses and organizations secure access to hundreds of on-chain ML algorithms as well as GNY’s proprietary data preparation. GNY’s goal is to transform the way groups share, collaborate, and analyze all kinds of data from financial records, climate change data, and vital public health information. Visit https://www.gny.io/a-new-collaborative-approach-to-tackle-covid-19-analysis/ to learn more about this milestone in decentralized data sharing and analysis.
Originally published: June 1, 2020