Derivative-Free Optimization for Population Dynamic Models
Quantifying populations in changing environments involves fitting highly non-linear and non-convex population dynamic models to distorted observations. Derivatives of the objective function with respect to parameters might be expensive to obtain, unreliable or unavailable.
The aim of this paper is to illustrate the use of derivative-free optimization for estimating parameters in continuous population dynamic models described by ordinary differential equations. A set of non-linear least squares problems is used to compare several solvers in terms of accuracy, computational costs and robustness. We also investigate criteria for a good optimization method which are specific to the type of objective function considered here. We see larger variations in the performances of the derivative-free methods when applied for parameter estimation in population dynamic models than observed for standard noisy benchmark problems.
KeywordsParameter Estimation Problem Direct Search Method Population Dynamic Model Mesh Adaptive Direct Search Pattern Search Method
Unable to display preview. Download preview PDF.
- 1.Abramson, M.A., Audet, C., Couture, G., Dennis Jr., J.E., Le Digabel, S., Tribes, C.: The NOMAD Project, https://www.gerad.ca/nomad/
- 3.Audet, C., Le Digabel, S., Tribes, C.: NOMAD user guide. Tech. Rep. G-2009-37, Les cahiers du GERAD (2009)Google Scholar
- 8.Fowler, K.R., Reese, J.P., Kees, C.E., Dennis Jr., J.E., Kelley, C.T., Miller, C.T., Audet, C., Booker, A.J., Couture, G., Darwin, R.W., Farthing, M.W., Finkel, D.E., Gablonsky, J.M., Gray, G., Kolda, T.G.: A comparison of derivative-free optimization methods for groundwater supply and hydraulic capture community problems. Adv. Water Resour. 31, 743–757 (2008)CrossRefGoogle Scholar
- 11.Le Digabel, S.: Algorithm 909: NOMAD: Nonlinear optimization with the MADS algorithm. ACM Transactions on Mathematical Software 37(4), 44:1–44:15 (2011)Google Scholar
- 14.Schaarschmidt, U., Subbey, S., Steihaug, T.: Application of slow-fast population dynamic models. Technical Report 409, Institute of Informatics, University of Bergen (2014)Google Scholar