Comparison of modern causality measures: PDC, TE, MIME and PMIME
- Datum
- 23.06.2014
- Zeit
- 10:30 - 12:00
- Sprecher
- Xiaogeng Wan
- Zugehörigkeit
- Imperial College London
- Sprache
- en
- Hauptthema
- Physik
- Andere Themen
- Physik
- Host
- CIDNET14 seminar talk
- Beschreibung
- We compare the three popular causality measures: transfer entropy (TE), Partial directed coherence (PDC) and conditional mutual information from mixed embedding (MIME). We use both theoretical maps such as Coupled Henon maps, Rossler-Lorenze systems and Macky-Glass systems and real data such as EEG data from ADD parients and financial data from real world market to compare the efficiency of those three measures, in terms of directionality, directness and computation time. Also, we construct analytical toy models for both Gaussian stantionary linear autoregressive vector processes and discrete processes (low dimensional Ising model), to compare the three measures. From those comparison, we present discussion on parameter choices and application tips.
Letztmalig verändert: 23.06.2014, 09:55:26
Veranstaltungsort
Max-Planck-Institut für Physik komplexer Systeme (Seminarroom 1)Nöthnitzer Straße3801187Dresden
- Telefon
- + 49 (0)351 871 0
- MPI-PKS
- Homepage
- http://www.mpipks-dresden.mpg.de
Veranstalter
Max-Planck-Institut für Physik komplexer SystemeNöthnitzer Straße3801187Dresden
- Telefon
- + 49 (0)351 871 0
- MPI-PKS
- Homepage
- http://www.mpipks-dresden.mpg.de
Legende
- Ausgründung/Transfer
- Bauing., Architektur
- Biologie
- Chemie
- Elektro- u. Informationstechnik
- für Schüler:innen
- Gesellschaft, Philos., Erzieh.
- Informatik
- Jura
- Maschinenwesen
- Materialien
- Mathematik
- Medizin
- Physik
- Psychologie
- Sprache, Literatur und Kultur
- Umwelt
- Verkehr
- Weiterbildung
- Willkommen
- Wirtschaft
