A Continuation Technique for Maximum Likelihood Estimators in Biological Models

• DOI: 10.1007/s11538-023-01200-0
• Publication Date: 2023-08-31T08:02:16Z
• Authors: Tyler Cassidy
• Keywords: Leverage (statistics); Experimental data; Continuation; Data set

Estimating model parameters is a crucial step in mathematical modelling and typically involves minimizing the disagreement between predictions experimental data. This calibration data can change throughout study, particularly if performed simultaneously with experiments, or during an on-going public health crisis as case of COVID-19 pandemic. Consequently, optimal parameter set, maximal likelihood estimator (MLE), function set. Here, we develop numerical technique to predict evolution MLE We show that, when considering perturbations from initial our approach significantly more computationally efficient that re-fitting while producing acceptable fits updated use continuation explicit functional relationship fit be used measure sensitivity then leverage this select similar information criteria, priori determine measurements which most sensitive, suggest additional experiment resolve uncertainty.