Competition challenged contestants to utilize data science toward firefighters’ safety on the scene during an emergency mission.
deepsense.io, a pure Apache Spark company providing code-free data manipulation and predictive analytics, today announced that Jan Lasek, deepsense.io Data Scientist, claimed victory in the high profile AAIA’15 Data Mining Competition: Tagging Firefighter Activities at a Fire Scene organized jointly by the University of Warsaw and the Main School of Fire Service, in Warsaw, Poland. It lasted over 3 months during which Jan beat 79 contestants, who submitted a total of 1,840 proposals.
deepsense.io machine learning team loves challenges and successfully participate in many machine learning and data science competitions. The contest won by Jan Lasek focused on creating a decision-support tool at an emergency scene. The goal was to develop a model that recognizes what activity a fireman is performing based on sensory data from his body movements and a collection of statistics monitoring his vital functions. The competition featured a cash award as well as an opportunity to present the winning solution at the 2015 Federated Conference on Computer Science and Information Systems.
“It was the first time I took part in such a competition. I like my job and the idea that data science can contribute to the safety of firemen. The realization that data science is growing in importance and given an opportunity to improve the lives of firemen inspired me to take part in this contest,” said deepsense.io’s Jan Lasek, a Data Scientist & Data Engineer. “Hopefully, the results of the competition will be used for better control over an emergency scene.”
The winning solution involved extensive feature engineering and an ensemble of random forest classifiers. Jan used R for statistical computation, data exploration and to test his approach. To find out more on his winning strategy, please read his blog post: Machine Learning for Greater Fire Scene Safety.
“I am proud to see Jan, a member of our machine learning team, have his skills put to the test and win such an inspiring competition. This is a demonstration of the high level of expertise among our professionals,” said Robert Bogucki, Chief Science Officer at deepsense.io. “This was a challenging, but rewarding problem to approach. Jan’s hard work, persistence and creativity during this competition show us how machine learning can contribute to greater safety at a fire scene.”
To learn more about the competition: AAIA ‘15
To learn more about deepsense.io Machine Learning services: BIG DATA SCIENCE SERVICES
Brand Manager at deepsense.io
deepsense.io trademarks at boilerplate