A Study Of Multi-Exponential Processes Analysis Methods

Gede , Pramudya Ananta (2012) A Study Of Multi-Exponential Processes Analysis Methods. Project Report. UTeM, Melaka, Malaysia. (Submitted)

PDF (Full Text) A_Study_Of_Multi-Exponential_Processes_Analysis_Methods_Gede_Pramudya_Ananta_QA372.G42_2012_DAH_EDIT.pdf - Submitted Version Restricted to Registered users only Download (10MB)

Abstract

There are numerous problems in sciences involving the resolution of experimental data as a linear combination of multi-exponential term in the form of n f(t) = LA;e-"''1 i=l in which the parameters amplitude Ai and rate constant A.iposses biological or physical significance. On the basis of data set DN = {(ti,yi):j = 1,2, ... ,N}, the task is to determine the parameters n and Fn = {(Ai, A.J: i = 1,2, ... , n & f(t) best fit the data DN}. It is assumed, form physical reasons, that the parameters are real positive. Four classes of methods were explored for solving the task.. In order to explore the relative advantages of a method over the others, review and observation were done. Sensitivity of the methods was also examined using several artificially simulated test data. The result of the exploration is a suggestion of using four-method sequential scheme and several facts about the application of the methods. Considering the limitation of the researchers, some significant works may be considered to be done in the future.

Actions (login required)

View Item