The algorithm developed by Prof. Evgeny Burnaev and Skoltech MSc student Vladislav Ishimtsev, took the third place in a competition devoted to online anomaly detection. This competition has been held for the first time this year as a part of the IEEE World Congress on Computational Intelligence.
Experts from the USA, India, Russia and other countries had to compete in accuracy of anomaly detection using the “Numenta Anomaly Benchmark” framework, consisting of a specialized testing system, developed by Numenta company, and a set of test problems, which includes more than 50 different real-world time-series with labels.
The competition took place from February till July 2016. In the category “Algorithms” Skoltech researchers from the Center for Computational Data-Intensive Science and Engineering (CDISE) prof. Evgeny Burnaev and MSc student Vladislav Ishimtsev took the third place.
Evgeny Burnaev: “I and my colleagues, CDISE master and PhD students, are working on methods of Machine Learning and their applications for industrial analytics. This competition provided to us a particularly valuable opportunity to test the algorithms we have developed and to compare their accuracy with accuracy of algorithms, developed by other participants.
Our result is a good achievement. In practice algorithms for anomaly detection are used to solve predictive maintenance problems. The main goal of predictive maintenance is to detect unwanted changes (anomalies) and to predict failures in complex engineering systems (e.g., engines, gas turbines, etc.) using data collected from sensors. E.g. we already have a positive experience in the practical application of such algorithms for solving predictive maintenance problems for various subsystems (e.g., auxiliary power unit, engine, etc.) of a passenger aircraft”.