Thesis supervisor: Zoltán Somogyvári
Location of studies (in Hungarian): Wigner Research Centre for Physics Abbreviation of location of studies: MTA
Description of the research topic:
From philosophers of ancient times to modern economists, researchers are engaged in revealing causal relations. Detecting causal relations between observed components without intervention is a challenging endeavor still today, while it is especially useful in order to understand complex systems such as the ecosystem of our planet or the human brain, where interventions are impossible or prohibited. We have developed the Dimensional Causality (DC) analysis method, using a Bayesian formulation applied to the theory of dynamical systems. The DC method is devised to detect and quantify the probability of all possible types of causal relationships between two time series: independence, direct or circular causal connection, and the existence of a hidden common cause (https://arxiv.org/abs/1808.10806). The planed PhD work has a double aim: To develop new and general mathematical data analysis tools, based on dynamical systems theory and to apply the developed methods to reveal causal interactions and networks in various fields of science, such as interactions between brain areas during normal and epileptic activity, to reveal spatio-temporal changes of causal connections of living species in artificial and natural ecosystems and to understand the effect of global warming on them. On methodological side, the planed work includes development of new causality analysis techniques between a continuous signal and a temporal point process, based on the previous results. This method has special interest in neuroscience, to reveal causal dependencies between extracellularly observed waves of synaptic currents and the spiking activity of the neurons. This would make possible to infer the actual spaio-temporal distribution of input currents on single neurons, based on multi-electrode measurements. We also aim to develop a method for hidden common cause reconstruction. While our recent Dimensional Causality method is able to reveal the existence of a hidden common cause, we are developing a new method which provides a topological map of the hidden common cause time series as well.
Required language skills: English Further requirements: Primarily, the planed work requires motivation and wide interest on multiple fields of science as well as theoretical and programming skills. Good communication skills are advantageous as the work will be done in national and international collaborat
Number of students who can be accepted: 1
Deadline for application: 2020-05-31
2024. IV. 17. ODT ülés Az ODT következő ülésére 2024. június 14-én, pénteken 10.00 órakor kerül sor a Semmelweis Egyetem Szenátusi termében (Bp. Üllői út 26. I. emelet).
Login
