arXiv subjects

Peter G. Taylor

Publications and source records attributed to Peter G. Taylor.

13 recordsLinked to original sources

arXiv2026

A multitype Markovian branching process with one-type population size dependence

Motivated by a within-host framework of immunity-modulated parasitic disease, we formulate a multitype Markovian branching process with reproductive parameters dependent on the number of individuals of a single type only. We impose a soft carrying capacity $K$ with respect to the controlling type, serving as a threshold between subcritical and supercritical dynamics. We prove that extinction occurs almost surely when the offspring batch of each non-controlling type includes an individual of the controlling type with positive probability, with the time to extinction possessing finite moments of all orders. We then construct a sequence of density-dependent processes indexed by $K$ and study scaling limits of the process using the canonical functional law of large numbers (FLLN) and central limit theorem, with a view towards characterising the emergence of metastable behaviour in the vicinity of asymptotically stable equilibria, or stable limit cycles of the FLLN.

math.PR↗

arXiv2025

Performance Evaluation of Small Call Centres in a Transient Regime

This paper addresses a fundamental and practically significant problem in call centre operations -- determining optimal call allocation policies that meet client service targets while minimising staffing costs. Motivated by a problem presented by an industry partner, we examine a real-world setting involving a relatively small call centre with hierarchical structure among agents. It is natural to model the operation of such a centre as a continuous-time Markov chain. To gain insight into the structure of optimal policies, we first (i) apply backward induction based on Bellman's equation to a finite-horizon discrete-time model, and (ii) derive stationary policies for an infinite-horizon continuous-time model with discounting. Subsequently, we evaluate the performance of these policies in the original finite-horizon continuous-time setting by computing the expected number of abandonments and the waiting time distributions of customers. This is achieved using first-step analysis combined with Laplace transform methods. The effectiveness of the proposed approach is illustrated through numerical examples.

math.PR↗

arXiv2025

Minimising Numbers of Losses and Abandonments in Small Call Centres Under a Transient Regime

We consider the problem of allocating customers to agents in small call centres so that transient performance indicators in terms of expected numbers of losses and abandonments are optimised. To gain insight into the general structure of optimal policies, we start by (i) using backward induction based upon Bellman's equation for a finite-horizon discrete-time model, and (ii) deriving stationary policies for an infinite-horizon continuous-time model with discounting. We then compare the performance of such policies in our original finite-horizon continuous-time model using a first-step analysis applied to the Laplace transforms of the relevant measures.

math.PR↗

arXiv2025

Superinfection and the hypnozoite reservoir for Plasmodium vivax: a multitype branching process approximation

Plasmodium vivax malaria is a mosquito-borne disease of significant public health importance. A defining feature of the within-host biology of P. vivax is the accrual of a hypnozoite reservoir, comprising a bank of quiescent parasites in the liver that are capable of causing relapsing blood-stage infections upon activation. Superinfection, characterised by composite blood-stage infections with parasites derived from multiple mosquito inoculation or hypnozoite activation events, is another important attribute. We have previously developed a stochastic epidemic model of P. vivax malaria, formulated as a Markov population process with countably infinitely-many types, that is adjusted for both hypnozoite accrual and blood-stage superinfection. Here, we construct a Markovian branching process with countably infinitely-many types to approximate the early stages of this epidemic model. With $P_M$ denoting the mosquito population size, we consider the limit $P_M \to \infty$ when the ratio of the mosquito and human populations is held fixed. We use a classical coupling argument to obtain a total variation bound of order $O(P_M^{2 κ- 1})$ that is valid until $o(P_M^κ)$ human-to-mosquito and mosquito-to-human transmission events have occurred, where $κ< 1/2$ is an arbitrary constant. We characterise the probability of global disease extinction under the branching process to approximate the probability of elimination, as opposed to sustained endemic transmission, when the epidemic model is initialised with low-level human and/or mosquito infection. We apply our model to two scenarios of epidemiological interest, namely the re-introduction of P. vivax malaria in a region where elimination has previously been achieved; and a mass drug administration campaign with population-wide depletion of the hypnozoite reservoir.

q-bio.PE↗

arXiv2023

A Stochastic Fluid Model Approach to the Stationary Distribution of the Maximum Priority Process

In traditional priority queues, we assume that every customer upon arrival has a fixed, class-dependent priority, and that a customer may not commence service if a customer with a higher priority is present in the queue. However, in situations where a performance target in terms of the tails of the class-dependent waiting time distributions has to be met, such models of priority queueing may not be satisfactory. In fact, there could be situations where high priority classes easily meet their performance target for the maximum waiting time, while lower classes do not. Here, we are interested in the stationary distribution at the times of commencement of service of this maximum priority process. Until now, there has been no explicit expression for this distribution. We construct a mapping of the maximum priority process to a tandem fluid queue, which enables us to find expressions for this stationary distribution. We derive the results for the stationary distribution of the maximum priority process at the times of the commencement of service.

math.PR↗

arXiv2023

Superinfection and the hypnozoite reservoir for Plasmodium vivax: a general framework

Malaria is a parasitic disease, transmitted by mosquito vectors. Plasmodium vivax presents particular challenges for disease control, in light of an undetectable reservoir of latent parasites (hypnozoites) within the host liver. Superinfection, which is driven by temporally proximate mosquito inoculation and/or hypnozoite activation events, is an important feature of P. vivax. Here, we present a model of hypnozoite accrual and superinfection for P. vivax. To couple host and vector dynamics, we construct a density-dependent Markov population process with countably many types, for which disease extinction is shown to occur almost surely. We also establish a functional law of large numbers, taking the form of an infinite-dimensional system of ordinary differential equations that can also be recovered under the hybrid approximation or a standard compartment modelling approach. Recognising that the subset of these equations that models the infection status of human hosts has precisely the same form as the Kolmogorov forward equations for a Markovian network of infinite server queues with an inhomogeneous batch arrival process, we use physical insight into the evolution of the latter to write down a time-dependent multivariate generating function for the solution. We use this characterisation to collapse the infinite-compartment model into a single integrodifferential equation (IDE) governing the intensity of mosquito-to-human transmission. Through a steady state analysis, we recover a threshold phenomenon for this IDE in terms of a bifurcation parameter $R_0$, with the disease-free equilibrium shown to be uniformly asymptotically stable if $R_0<1$ and an endemic equilibrium solution emerging if $R_0>1$. Our work provides a theoretical basis to explore the epidemiology of P. vivax, and introduces a general strategy for constructing tractable population-level models of malarial superinfection.

q-bio.PE↗

arXiv2023

Open networks of infinite server queues with non-homogeneous multivariate batch Poisson arrivals

In this paper, we consider the occupancy distribution for an open network of infinite server queues with multivariate batch arrivals following a non-homogeneous Poisson process, and general service time distributions. We derive a probability generating function for the transient occupancy distribution of the network, and prove that it is necessary and sufficient for ergodicity that the expected occupancy time for each batch be finite. Further, we recover recurrence relations for the transient probability mass function formulated in terms of a distribution obtained by compounding the batch size with a multinomial distribution.

math.PR↗

arXiv2022

A hybrid transmission model for Plasmodium vivax accounting for superinfection, immunity and the hypnozoite reservoir

Malaria is a vector-borne disease that exacts a grave toll in the Global South. The epidemiology of Plasmodium vivax, the most geographically expansive agent of human malaria, is characterised by the accrual of a reservoir of dormant parasites known as hypnozoites. Relapses, arising from hypnozoite activation events, comprise the majority of the blood-stage infection burden, with implications for the acquisition of immunity and the distribution of superinfection. Here, we construct a hybrid transmission model for P. vivax that concurrently accounts for the accrual of the hypnozoite reservoir, (blood-stage) superinfection and the acquisition of immunity. We begin by analytically characterising within-host dynamics as a function of mosquito-to-human transmission intensity, extending our previous model (comprising an open network of infinite server queues) to capture a discretised immunity level. To model transmission-blocking and antidisease immunity, we allow for geometric decay in the respective probabilities of successful human-to-mosquito transmission and symptomatic blood-stage infection as a function of this immunity level. Under a hybrid approximation -- whereby probabilistic within-host distributions are cast as expected population-level proportions -- we couple host and vector dynamics to recover a deterministic compartmental model in line with Ross-Macdonald theory. We then perform a steady-state analysis for this compartmental model, informed by the (analytic) distributions derived at the within-host level. To characterise transient dynamics, we derive a reduced system of integrodifferential equations (IDEs), likewise informed by our within-host queueing network, allowing us to recover population-level distributions for various quantities of epidemiological interest. Our model provides insights into important, but poorly understood, epidemiological features of P. vivax.

q-bio.PE↗

arXiv2021

An activation-clearance model for Plasmodium vivax malaria

Malaria is an infectious disease with an immense global health burden. Plasmodium vivax is the most geographically widespread species of malaria. Relapsing infections, caused by the activation of liver-stage parasites known as hypnozoites, are a critical feature of the epidemiology of Plasmodium vivax. Hypnozoites remain dormant in the liver for weeks or months after inoculation, but cause relapsing infections upon activation. Here, we introduce a dynamic probability model of the activation-clearance process governing both potential relapses and the size of the hypnozoite reservoir. We begin by modelling activation-clearance dynamics for a single hypnozoite using a continuous-time Markov chain. We then extend our analysis to consider activation-clearance dynamics for a single mosquito bite, which can simultaneously establish multiple hypnozoites, under the assumption of independent hypnozoite behaviour. We derive analytic expressions for the time to first relapse and the time to hypnozoite clearance for mosquito bites establishing variable numbers of hypnozoites, both of which are quantities of epidemiological significance. Our results extend those in the literature, which were limited due to an assumption of non-independence. Our within-host model can be embedded readily in multi-scale models and epidemiological frameworks, with analytic solutions increasing the tractability of statistical inference and analysis. Our work therefore provides a foundation for further work on immune development and epidemiological-scale analysis, both of which are important for achieving the goal of malaria elimination.

q-bio.PE↗

arXiv2020

A Restless Bandit Model for Resource Allocation, Competition and Reservation

We study a resource allocation problem with varying requests, and with resources of limited capacity shared by multiple requests. It is modeled as a set of heterogeneous Restless Multi-Armed Bandit Problems (RMABPs) connected by constraints imposed by resource capacity. Following Whittle's relaxation idea and Weber and Weiss' asymptotic optimality proof, we propose a simple policy and prove it to be asymptotically optimal in a regime where both arrival rates and capacities increase. We provide a simple sufficient condition for asymptotic optimality of the policy, and in complete generality propose a method that generates a set of candidate policies for which asymptotic optimality can be checked. The effectiveness of these results is demonstrated by numerical experiments. To the best of our knowledge, this is the first work providing asymptotic optimality results for such a resource allocation problem and such a combination of multiple RMABPs.

math.OC↗

arXiv2018

Why is Kemeny's constant a constant?

In their 1960 book on finite Markov chains, Kemeny and Snell established that a certain sum is invariant. The value of this sum has become known as {\it Kemeny's constant}. Various proofs have been given over time, some more technical than others. We give here a very simple physical justification, which extends without a hitch to continuous-time Markov chains on a finite state space. For Markov chains with denumerably infinite state space, the constant may be infinite and even if it is finite, there is no guarantee that the physical argument will hold. We show that the physical interpretation does go through for the special case of a birth-and-death process with a finite value of Kemeny's constant. Keywords: Kemeny's constant; discrete-time Markov chains; continuous-time Markov chains; passage times; deviation matrix.

math.PR↗

arXiv2015

Calculating optimal limits for transacting credit card customers

We present a model of credit card profitability, assuming that the card-holder always pays the full outstanding balance. The motivation for the model is to calculate an optimal credit limit, which requires an expression for the expected outstanding balance. We derive its Laplace transform, assuming that purchases are made according to a marked point process and that there is a simplified balance control policy in place to prevent the credit limit being exceeded. We calculate optimal limits for a compound Poisson process example and show that the optimal limit scales with the distribution of the purchasing process and that the probability of exceeding the optimal limit remains constant. We establish a connection with the classic newsvendor model and use this to calculate bounds on the optimal limit for a more complicated balance control policy. Finally, we apply our model to real credit card purchase data.

math.OC↗

arXiv2015

Bitcoin Blockchain Dynamics: the Selfish-Mine Strategy in the Presence of Propagation Delay

In the context of the `selfish-mine' strategy proposed by Eyal and Sirer, we study the effect of propagation delay on the evolution of the Bitcoin blockchain. First, we use a simplified Markov model that tracks the contrasting states of belief about the blockchain of a small pool of miners and the `rest of the community' to establish that the use of block-hiding strategies, such as selfish-mine, causes the rate of production of orphan blocks to increase. Then we use a spatial Poisson process model to study values of Eyal and Sirer's parameter $γ$, which denotes the proportion of the honest community that mine on a previously-secret block released by the pool in response to the mining of a block by the honest community. Finally, we use discrete-event simulation to study the behaviour of a network of Bitcoin miners, a proportion of which is colluding in using the selfish-mine strategy, under the assumption that there is a propagation delay in the communication of information between miners.

cs.CR↗