LAUSANNE, Switzerland, June 20, 2019 /PRNewswire-PRWeb/ -- Datapred, an innovator in machine learning software, today announced that it has won the time series modeling challenge organized by Airbus.
During the award ceremony in Toulouse on June 13, Adam Bonnifield, VP Artificial Intelligence at Airbus, commented that "real-time situation awareness and responsiveness is critical for companies like Airbus." He added that "the ability quickly to build, test and deploy industry-strength applications that make sense of event streams while avoiding the pitfalls of time series modeling is a blessing for our data scientists."
The goal of the challenge was to detect anomalies in 1,000 sequences of in-flight data from 90 sensors per commercial aircraft. One notable difficulty was the absence of a set of known anomalies ("unsupervised learning"). 219 teams (companies, university laboratories and independent data scientists) registered worldwide, with Airbus receiving 38 final submissions.
"We are very proud of this external validation for the Datapred modeling engine," said Nicolas Mahler, Director of Datapred. "With the proliferation of connected sensors, time series are invading industrial data sets. We believe that building robust continuous intelligence applications requires capabilities that open-source machine learning libraries and all-purpose machine learning platforms don't provide."
Datapred's modeling engine powers the company's machine learning for direct material procurement software. It is also used by enterprises for asset performance monitoring and supply chain optimization. It is available to data scientists on the AWS and Microsoft Azure marketplaces.
About Datapred
Founded in 2015 and based in Lausanne, Switzerland, Datapred provides a modeling engine for continuous intelligence that streamlines everything that is specific to machine learning for time series, at every stage of the machine learning pipeline. Benefits include greater regression and classification performance and a 10x acceleration of the garage-to-factory cycle.
Additional Information
For additional information on Datapred, please visit https://www.datapred.com/contact.
SOURCE Datapred
Share this article