ebook img

Tackling the Inverse Problem for Non-Autonomous Systems: Application to the Life Sciences PDF

145 Pages·2014·2.48 MB·English
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Tackling the Inverse Problem for Non-Autonomous Systems: Application to the Life Sciences

Springer Theses Recognizing Outstanding Ph.D. Research Tomislav Stankovski Tackling the Inverse Problem for Non-Autonomous Systems Application to the Life Sciences Springer Theses Recognizing Outstanding Ph.D. Research For furthervolumes: http://www.springer.com/series/8790 Aims and Scope The series ‘‘Springer Theses’’ brings together a selection of the very best Ph.D. theses from around the world and across the physical sciences. Nominated and endorsed by two recognized specialists, each published volume has been selected for its scientific excellence and the high impact of its contents for the pertinent fieldofresearch.Forgreateraccessibilitytonon-specialists,thepublishedversions includeanextendedintroduction,aswellasaforewordbythestudent’ssupervisor explaining the special relevance of the work for the field. As a whole, the series will provide a valuable resource both for newcomers to the research fields described, and for other scientists seeking detailed background information on specialquestions.Finally,itprovidesanaccrediteddocumentationofthevaluable contributions made by today’s younger generation of scientists. Theses are accepted into the series by invited nomination only and must fulfill all of the following criteria • They must be written in good English. • ThetopicshouldfallwithintheconfinesofChemistry,Physics,EarthSciences, Engineering andrelatedinterdisciplinaryfieldssuchasMaterials, Nanoscience, Chemical Engineering, Complex Systems and Biophysics. • The work reported in the thesis must represent a significant scientific advance. • Ifthethesisincludespreviouslypublishedmaterial,permissiontoreproducethis must be gained from the respective copyright holder. • They must have been examined and passed during the 12 months prior to nomination. • Each thesis should include a foreword by the supervisor outlining the signifi- cance of its content. • The theses should have a clearly defined structure including an introduction accessible to scientists not expert in that particular field. Tomislav Stankovski Tackling the Inverse Problem for Non-Autonomous Systems Application to the Life Sciences Doctoral Thesis accepted by Lancaster University, UK 123 Author Supervisor Dr. TomislavStankovski Prof.Aneta Stefanovska Department of Physics Department of Physics Lancaster University Lancaster University Lancaster Lancaster UK UK ISSN 2190-5053 ISSN 2190-5061 (electronic) ISBN 978-3-319-00752-6 ISBN 978-3-319-00753-3 (eBook) DOI 10.1007/978-3-319-00753-3 SpringerChamHeidelbergNewYorkDordrechtLondon LibraryofCongressControlNumber:2013940308 (cid:2)SpringerInternationalPublishingSwitzerland2014 Thisworkissubjecttocopyright.AllrightsarereservedbythePublisher,whetherthewholeorpartof the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,broadcasting,reproductiononmicrofilmsorinanyotherphysicalway,andtransmissionor informationstorageandretrieval,electronicadaptation,computersoftware,orbysimilarordissimilar methodology now known or hereafter developed. Exempted from this legal reservation are brief excerpts in connection with reviews or scholarly analysis or material supplied specifically for the purposeofbeingenteredandexecutedonacomputersystem,forexclusiveusebythepurchaserofthe work. Duplication of this publication or parts thereof is permitted only under the provisions of theCopyright Law of the Publisher’s location, in its current version, and permission for use must always be obtained from Springer. Permissions for use may be obtained through RightsLink at the CopyrightClearanceCenter.ViolationsareliabletoprosecutionundertherespectiveCopyrightLaw. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publicationdoesnotimply,evenintheabsenceofaspecificstatement,thatsuchnamesareexempt fromtherelevantprotectivelawsandregulationsandthereforefreeforgeneraluse. While the advice and information in this book are believed to be true and accurate at the date of publication,neithertheauthorsnortheeditorsnorthepublishercanacceptanylegalresponsibilityfor anyerrorsoromissionsthatmaybemade.Thepublishermakesnowarranty,expressorimplied,with respecttothematerialcontainedherein. Printedonacid-freepaper SpringerispartofSpringerScience+BusinessMedia(www.springer.com) Supervisor’s Foreword In simple terms science might be defined as the systematic observation of nature andofhownaturalprocessesevolveintimeandspace.Suchprocessescaninclude thebeatingoftheheartofalivinghumanorothermammal,themovementsofthe planets, or simply how human society works. In all these cases, physics tries to generatea model based on data collectedovertime—a time interval thatdepends on how fast the processes occur, which may in some cases happen over centuries and in others over seconds or microseconds. Science attempts to develop the models that can most comprehensively link the causes and consequences of the processes in question. One of the most frequently-used approaches, and arguably the most useful one, is the Bayesian approach. It is based on Bayes’ theorem, whichiscentraltotheinverseproblemapproachanddynamicalinference,seeking toanswerthequestion:givenaseriesofdataresultingfromobservations,whatcan we deduce about the nature of the system or the process that generated that data? Thomas Bayes was fortunate that his friend Richard Price significantly edited and updated his work and read it to the Royal Society in 1763, a year after his death. It was published in the Philosophical Transactions of the Royal Society of London the following year. The ideas gained limited exposure until they were independentlyrediscoveredandfurtherdevelopedbyLaplace,whofirstpublished the modern formulation in his 1812 Théorie analytique des probabilités. The classical approach to statistics defines the probability of an event as ‘‘The numberoftimestheeventoccursoverthetotalnumberoftrials,inthelimitofan infinite series of equiprobable repetitions’’, which is quite limited. Many of these limitations can be avoided and paradoxes resolved by taking a Bayesian stance about probabilities. Bayes defines probability as: The probability of any event is the ratio between the value at which an expectation depending on the happening of the event ought to be computed, and the value of the thing expected upon its happening. HoweverevenBayeshimselfmightnothaveembracedthebroadinterpretation now called Bayesian. It is difficult to assess Bayes’ philosophical views on probability, since his work does not go into questions of interpretation. TodayBayesianprobabilityisthenamegiventoseveralrelatedinterpretations of probability. To evaluate the probability of a hypothesis, the Bayesian v vi Supervisor’sForeword probability specifies some prior probability, which is then updated in the light of new, relevant data. ‘‘Bayesian’’hasbeen used in this sense since its rebirth in the 1950s. Advances in computer technology have allowed scientists from many disciplines to extend and use this approach in diverse fields. Sir Harold Jeffreys , whowrotethebookTheoryofProbability,whichfirstappearedin1939,playedan important role in the revival of the Bayesian view of probability. He wrote that Bayes’ theorem ‘‘is to the theory of probability what Pythagoras’s theorem is to geometry’’. Recently,wehavefacedhugedevelopmentsinboththetheoryandapplications of Bayes theorem. The applications span practically every aspect of science, including particle physics, astrophysics, cosmology, geophysics, and medical and biological physics. The work by Tomislav Stankovski deals with one particular aspect of Bayesian inference: it focuses on inference of the dynamical properties of oscillatory systems. The work is based on the development of Feynman’s path integral whose central idea is that, for the motion of a particle between two points in space, all possible connecting trajectories should be considered and a probability amplitude assignedtoeachoneofthem.Thispathintegralgiveslikelihoodofobservationof thedynamicaltrajectoryforagivensetofdistributionsofmodelparameters.Once theactualdynamicaltrajectoryismeasuredintheexperimentthedistributionsfor the set of model parameters can be improved using Bayes theorem. The other important ingredient is based on recent developments in the under- standing of the nonlinear dynamics of oscillatory processes, and the theory of synchronization in particular. The main objective of Tomislav’s thesis, now pre- sentedasabook,wastostudy,detectandunderstandingreaterdetailtheeffectof external dynamical influences on interacting selfsustained oscillators. The work was motivated by problems associated with biological systems and thecardiovascularsysteminparticular.Notsurprisingly,lifecannotbeunderstood properly without a self-consistent theory of non-autonomous systems and associ- ated methods to infer such dynamical characteristics from measured data. How- ever,non-autonomicityischaracteristicofmanysystems.Inreality,practicallyall systemsarenon-autonomous,thoughsomecansafelybesimplifiedandstudiedas closed or isolated. Hence, I can envisage very wide applicability of the method proposed in this book. Despite a long-standing general awareness of non-autonomous systems, rela- tively little has been done in this field. Mathematics has included mainly the process and the skew product flow formalism, while the physics approach to date has mainly been to reduce the dynamics to being autonomous by adding an extra dimension for the time-dependence. Supervisor’sForeword vii Tomislav’s work is one of the first systematic approaches to the treatment of non-autonomoussystemsfromaphysicsperspectiveand,inparticular,todoingso inaninverseapproach. Asamultitudeofdataarewidelyavailable today,wecan expect highly significant advances in many fields to be facilitated by the results presented in this book. Lancaster, May 2013 Prof. Aneta Stefanovska Acknowledgments First and foremost I offer my sincerest gratitude to my supervisor, Prof. Aneta Stefanovska, who introduced me to the world of nonlinear biomedical physics. I am very grateful for her unreserved support throughout my thesis, guiding me with her knowledge, experience and enthusiasm. IwishtosincerelythankProf.PeterV.E.McClintockforhissupport,guidance and willingness to share his great experience and wisdom. IespeciallywishtothankDr.AndreaDuggento,withwhomIhavecollaborated on much of the work presented. He was great tutor, colleague and friend, and it was a great pleasure to work with him. IwishtothankProf.DwainEckbergforthecollaborationandforprovidingthe data for the physiological study. I appreciate greatly the valuable discussion with Dr. Martin Rasmussen, Prof. Peter Kloeden and Dr. Martin Wechselberger regardingthestudyofnon-autonomoussystems.IwishalsotothankProf.Metodija Kamilovskiforhisadvicesonanaloguesimulation. A special gratitude for both scientific assistance and kind friendship goes to manyofmycolleaguesDr.AlanBernjak,Dr.AlirezaBahraminasab,Dr.Lawrence Sheppard, Dr. David Kenwright, Dr. Rodrigue Tindjong, Dr. Martin Horvat, Dr. Uchechukwu Vincent, Adam Bradbury, Dmytro Iatsenko. For valuable discussionandimprovingthewrittenthesis,Ialsowishtothankmycolleaguesand friendsPhilClemson,GemmaLancaster,PhilipStephens,YvannStephens,Rachel Sparks,Spase Petkoski andDr.YevhenSuprunenko. Iamespecially grateful tomyfriends andfamily, for their encouragementand support. I thank my father for inspiring me, and my mother for giving me the mental strength to cope with the challenges throughout my studies. Last,butbynomeansleast,Iwishtoexpressmygratitudetomybelovedwife. Her unreserved and unconditional support, understanding and encouragement helpedmetogetthroughoverthelast4years.Iwillneverbegratefulenoughtoher. ix Abstract The common assumption that a dynamical system found in nature can be considered as isolated and autonomous is frequently a poor approximation. In reality, there are always external influences, and these are often too strong to ignore. In the case of an interacting oscillatory systems, they may, e.g. modify theirnaturalfrequenciesorcouplingamplitudes.Themainobjectiveofthisthesis istostudy,detectandunderstandingreaterdetailtheeffectofexternaldynamical influences on interacting self-sustained oscillators. Theoretical framework for the analysis of synchronization between non- autonomousoscillatingsystemsisdiscussed.Multiple-scaleanalysisisappliedon a phase oscillators model with slowly varying frequency. This analysis revealed the analytic form of the synchronization state with respect to slow and fast time- variations. Limit-cycle oscillators are used to study amplitude dynamics and to investigate synchronization transitions, which occur in the bifurcation points where the equilibrium solution for the phase difference and amplitudes changes their stability. Bifurcation diagrams as functions of coupling parameters are also constructed. In a case of non-autonomous interacting oscillators, the phase difference varies dynamically, the external influences can be the cause for synchronization transitions between different synchronization orders, and lag synchronization is hardly achievable. It is also demonstrated that the time- variations of the form of the coupling function alone can be the cause for synchronization transitions. A method is introduced for analysis of interactions between time-dependent coupled oscillators, based on the signals they generate. It distinguishes unsynchronized dynamics from noise-induced phase slips, and enables the evolution of the coupling functions and other parameters to be followed. The technique is based on Bayesian inference of the time-evolving parameters, achieved by shaping the prior densities to incorporate knowledge of previous samples.Thedynamicscanbeinferredfromphasevariables,inwhichcaseafinite number of Fourier base functions are used, or from state variables exploiting the model state base functions. The latter is used for detection of generalized synchronization. The method is tested numerically and applied to reveal and quantify the time-varying nature of synchronization, directionality and coupling functionsfromcardiorespiratoryandanalogue signals.Itisfound that,incontrast xi

Description:
This thesis presents a new method for following evolving interactions between coupled oscillatory systems of the kind that abound in nature. Examples range from the subcellular level, to ecosystems, through climate dynamics, to the movements of planets and stars. Such systems mutually interact, adju
See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.