arXiv subjects

Simon D. W. Frost

Publications and source records attributed to Simon D. W. Frost.

3 recordsLinked to original sources

arXiv2025

Exploring epidemic control policies using nonlinear programming and mathematical models

Optimal control theory in epidemiology has been used to establish the most effective intervention strategies for managing and mitigating the spread of infectious diseases while considering constraints and costs. Using Pontryagin's Maximum Principle, indirect methods provide necessary optimality conditions by transforming the control problem into a two-point boundary value problem. However, these approaches are often sensitive to initial guesses and can be computationally challenging, especially when dealing with complex constraints. In contrast, direct methods, which discretise the optimal control problem into a nonlinear programming (NLP) formulation, could offer robust, adaptable solutions for real-time decision-making. Despite their potential, the widespread adoption of these techniques has been limited, which may be due to restricted access to specialised software, perceived high costs, or unfamiliarity with these methods. This study investigates the feasibility, robustness, and potential of direct optimal control methods using nonlinear programming solvers on compartmental models described by ordinary differential equations to determine the best application of various interventions, including non-pharmaceutical interventions and vaccination strategies. Through case studies, we demonstrate the use of NLP solvers to determine the optimal application of interventions based on single objectives, such as minimising total infections, "flattening the curve", or reducing peak infection levels, as well as multi-objective optimisation to achieve the best combination of interventions. While indirect methods provide useful theoretical insights, direct approaches may be a better fit for the fast-evolving challenges of real-world epidemiology.

physics.soc-ph↗

arXiv2015

Modeling and Analysing Respondent Driven Sampling as a Counting Process

Respondent-driven sampling (RDS) is an approach to sampling design and analysis which utilizes the networks of social relationships that connect members of the target population, using chain-referral methods to facilitate sampling. RDS typically leads to biased sampling, favoring participants with many acquaintances. Naive estimates, such as the sample average, which are uncorrected for the sampling bias, will themselves be biased. To compensate for this bias, current methodology suggests inverse-degree weighting, where the "degree" is the number of acquaintances. This stems from the fundamental RDS assumption that the probability of sampling an individual is proportional to their degree. Since this assumption is tenuous at best, we propose to harness the additional information encapsulated in the time of recruitment, into a model-based inference framework for RDS. This information is typically collected by researchers, but ignored. We adapt methods developed for inference in epidemic processes to estimate the population size, degree counts and frequencies. While providing valuable information in themselves, these quantities ultimately serve to debias other estimators, such a disease's prevalence. A fundamental advantage of our approach is that, being model-based, it makes all assumptions of the data-generating process explicit. This enables verification of the assumptions, maximum likelihood estimation, extension with covariates, and model selection. We develop asymptotic theory, proving consistency and asymptotic normality properties. We further compare these estimators to the standard inverse-degree weighting through simulations, and using real-world data. In both cases we find our estimators to outperform current methods. The likelihood problem in the model we present is convex, and thus efficiently solvable. We implement these estimators in an R package, chords, available on CRAN.

stat.ME↗

arXiv2011

Analysis of a capture-recapture estimator for the size of populations with heterogenous catchability, and its evaluation on RDS data from rural Uganda

In this paper, we consider capture-recapture experiments with heterogenous catchability. In the setting we consider, the widespread Huggins-Alho estimator is not very suitable and we introduce and study a new generalized Horvitz-Thompson estimator. Our motivation is Respondent Driven Sampling (RDS), a prime example for such a setting where the capture probability is dependent on both the unknown population size as well as on an observable covariate, the network degree of an individual, due to peer recruitment. After discussing the theoretical properties of the new estimator, with full details given in the appendix, we evaluate it on various empirical and simulated data-sets, focusing on an RDS survey in a population in rural Uganda in which the population size is known a priori. The results thus obtained demonstrate that the adjusted estimator is less biased than the naive Lincoln-Petersen estimator.

stat.AP↗