Democritus University of Thrace (DUTH)
Μελέτη στοχαστικών συστημάτων με χρήση μη γραμμικών μοντέλων: εφαρμογές σε προβλήματα νευροφυσιολογίας
Abstract
dc:descriptionIn the present doctoral thesis was initially examined the cross-correlation between the events of two or three point processes. For this aim, the estimators of the product and cumulant densities of the increments of stationary point processes were defined and studied. These estimates are based on statistical functions from the time and the frequency domain. Then, we defined parameters in the time domain, suitable for the examination of the cross-correlation between two (product ratio) or three stationary point processes (conditional intensities). When the hypothesis of independence of point processes is rejected, these parameters allow evaluating the strength of association for various values of lagtime between the events. The statistical inference for the product ratio is based on spectral analysis and spectral densities of higher order (up to forth), which provide improved results using convergence factors, since the statistical inference for the conditional intensities is based on techniques from the time domain. In each case it is achieved the construction of confidence intervals with constant limits for each value of lagtime, applying transformations of our estimates, as the transformation of the square root. The results from the two approaches are equivalent. Nevertheless spectral analysis techniques allow making better smoothing in our estimates. Moreover, linear but also not linear point process stochastic models with one or two inputs and one output are studied. The above theoretical findings are applied for the study of the muscle spindle, a complex element of the neuromuscular system of mammals. Specifically, the excitatory effect of the gamma motoneuron, the inhibitory effect of the alpha motoneuron and their simultaneous activation on the mechanical properties of the muscle spindle is studied. The statistical elaboration and the graphical representation of the data sets with R, the well-known the statistical package R and the Fortran programming language.
Degree
thesis:*- Grantor dc:publisher
- Democritus University of Thrace (DUTH)
- Year dc:date
- 2009
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Karavasilis, Georgios
- Καραβασίλης, Γεώργιος
Subjects
dc:subject × 20- Στάσιμες σημειακές ανελίξεις
- Ανάλυση στο πεδίο του χρόνου
- Συναρτήσεις γινομένου και συσσώρευσης
- Ανάλυση στο πεδίο της συχνότητας
- Συναρτήσεις φασματικών πυκνοτήτων
- Στοχαστικά συστήματα
- Νευρομυϊκό σύστημα
- Στατιστικό λογισμικό R
- Stationary point processes
- Analysis in the time domain
- Product and cumulant densities
- Analysis in the frequency domain
- Spectral density functions
- Stochastic systems
- Neuromuscular system
- Statistical software R
- Επιστήμες Μηχανικού και Τεχνολογία
- Επιστήμη Ηλεκτρολόγου Μηχανικού, Ηλεκτρονικού Μηχανικού, Μηχανικού Η/Υ
- Engineering and Technology
- Electrical Engineering, Electronic Engineering, Information Engineering
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/18508