| Autor: |
Cajayon, Raquel C., Lucilo, Jayson A., Pilar-Arceo, Carlene P. C., Mendoza, Eduardo R. |
| Předmět: |
|
| Zdroj: |
Philippine Journal of Science; Mar2020, Vol. 149 Issue 1, p63-78, 25p |
| Abstrakt: |
Parameter estimation for models of biochemical systems is computationally expensive due to the nonlinearity and high dimensionality of the coupled systems of ordinary differential equations underlying the models. Hence, it is important to apply novel methods to the problem and evaluate their performance. We consider the Bat algorithm (BA) and the Firefly algorithm (FA) with respect to parameter estimation of S-system models. Using three S-systems of increasing complexity from the MADMan benchmarking framework, we assess and compare the relative performance of the two algorithms relative to various data sets, initial conditions, and noise levels. Simulation results show that both algorithms can be effectively used in estimating parameters of the S-system models. In particular, in all three S-systems used, the FA performed better than the BA based on the final cost function values and relative estimate errors. The introduction of noise to data significantly affected the convergence of both algorithms. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
