Kovács, Edith: High-dimensional data-driven modeling of …

Előadás címe: High-dimensional data-driven modeling of complex systems

Helyszín: Online, https://meet.google.com/ete-gaku-aex

Időpont: 2020.06.29., 15:55 – 16:15

Kivonat: In this talk we present the research we are conducting as a small research group in the framework of the HU-MATHS-IN research group, Statistics and Mathematical Modeling Consulting Group (BME) supported by the Hungarian Service Network for Mathematics in Industry and Innovations. Our work revolves around high-dimensional complex systems. The advancement in computing and storage capabilities of modern computational systems have made it possible to analyze high-dimensional complex systems, which opens up new research directions. Furthermore, high dimensionality also brings new challenges to the machine learning and data science communities. Modelling and analyzing high dimensional data is a major challenge in various domains. We will present important mathematical challenges with respect to high-dimensional complex systems together with real-world and industrial applications. One of the most important achievements is the copula-based anomaly detection system that we developed in cooperation with our industrial partner, NOKIA-Bell Labs. Our innovation was named as the most significant innovation at the Budapest University of Technology and Economics in 2019. Among the multiple recent results of our group, in the present talk we put emphasize on new discoveries on copula-based high dimensional data modeling.