Telcs, András
(Wigner RCP, BME SZIT)
Causal
discovery
From philosophers of ancient times to modern ones scientist of different fields as physicists economists, biologists engaged in revealing causal relations. The most challenging problem is inferring the type of the causal relationship: whether it is uni- or bi-directional or only apparent – implied by a hidden common cause only. Modern technology provides us tools to record data from complex systems such as the ecosystem of our planet or the human brain, but understanding their functioning needs detection and distinction of causal relationships of the system components without interventions given that in many cases experiment, intervention is not possible. There are two major class of systems to be investigate, deterministic and stochastic dynamic systems. Recently we proposed a method to detect and identify causal relation between deterministic systems. The method is based on the investigation of the attractor of the corresponding systems. Unfortunately, if the observation are noisy, we immediately face to a nondeterministic but stochastic system (not to speak about truly stochastic systems) and the method works but only on heuristic grounds. This lecture presents a new approach which provide unified treatment of the causal discovery between pairs of deterministic or stochastic dynamic systems We hope that the new method can be successfully applied to the analysis of EEG (electroencephalographic) data, climate data, plasma convection and even stock prices, in practice to root cause detection in large plant control systems’ data.
The talk is held in English!
Az előadás
nyelve angol!
Date: Sep 22, Tuesday 4:15pm
Place: BME, Building „Q”, Room
QBF13