Statistical analysis of three fourth-order ordinary differential equation solvers

Simple item page
dc.contributor.committeeChairMartin, Clyde F.
dc.contributor.committeeMemberSurles, James
dc.contributor.committeeMemberNeusel, Mara D.
dc.creatorHe, Bo
dc.date.accessioned2016-11-14T23:14:49Z
dc.date.available2012-06-01T16:36:35Z
dc.date.available2016-11-14T23:14:49Z
dc.date.issued2007-12
dc.degree.departmentStatistics
dc.description.abstractWe develop an autoregressive integrated moving average model (ARIMA) to study the statistical behavior of the numerical error generated from three fourth-order ODE Solvers: Milne's method, Adams-Bashforth method and a new method which randomly switches between Milne and Adams-Bashforth methods. With the actual error data based on three differential equations we desire to identify an ARIMA model to each data series. Results show that some of data series can be described by ARIMA models but others can not. Based on the mathematical form of the error data, other statistical models should be applied in the future. Finally we assess the multivariate normality of sample mean vectors which are generated by the switching method as an application of the multivariate central limit theorem.
dc.format.mimetypeapplication/pdf
dc.identifier.urihttp://hdl.handle.net/2346/12657
dc.language.isoeng
dc.rights.availabilityUnrestricted.
dc.subjectNumerical methods
dc.subjectMultivariate normality
dc.subjectAutoregressive integrated moving average model (ARIMA)
dc.titleStatistical analysis of three fourth-order ordinary differential equation solvers
dc.typeThesis

Files

Collections

Texas Tech University