Estimation for general birth-death processes

DOI10.1080/01621459.2013.866565zbMATH Open1367.62245arXiv1111.4954OpenAlexW2096272023WikidataQ42058110 ScholiaQ42058110MaRDI QIDQ4975412FDOQ4975412

Authors: Forrest W. Crawford, Vladimir N. Minin, Marc A. Suchard 

Publication date: 4 August 2017

Published in: Journal of the American Statistical Association (Search for Journal in Brave)

Abstract: Birth-death processes (BDPs) are continuous-time Markov chains that track the number of "particles" in a system over time. While widely used in population biology, genetics and ecology, statistical inference of the instantaneous particle birth and death rates remains largely limited to restrictive linear BDPs in which per-particle birth and death rates are constant. Researchers often observe the number of particles at discrete times, necessitating data augmentation procedures such as expectation-maximization (EM) to find maximum likelihood estimates. The E-step in the EM algorithm is available in closed-form for some linear BDPs, but otherwise previous work has resorted to approximation or simulation. Remarkably, the E-step conditional expectations can also be expressed as convolutions of computable transition probabilities for any general BDP with arbitrary rates. This important observation, along with a convenient continued fraction representation of the Laplace transforms of the transition probabilities, allows novel and efficient computation of the conditional expectations for all BDPs, eliminating the need for approximation or costly simulation. We use this insight to derive EM algorithms that yield maximum likelihood estimation for general BDPs characterized by various rate models, including generalized linear models. We show that our Laplace convolution technique outperforms competing methods when available and demonstrate a technique to accelerate EM algorithm convergence. Finally, we validate our approach using synthetic data and then apply our methods to estimation of mutation parameters in microsatellite evolution.

Full work available at URL: https://arxiv.org/abs/1111.4954

Recommendations

• Parameter estimation for discretely observed linear birth‐and‐death processes
• Estimation for partially observed birth–death processes
• Birth/birth-death processes and their computable transition probabilities with biological applications
• Statistical estimation of birth-and-death processes under incomplete observations
• The linear birth-death process: an inferential retrospective

zbMATH Keywords

EM algorithmmaximum likelihood estimationcontinuous-time Markov chainMM algorithmmicrosatellite evolution

Mathematics Subject Classification ID

Markov processes: estimation; hidden Markov models (62M05) Applications of statistics to biology and medical sciences; meta analysis (62P10) Continuous-time Markov processes on discrete state spaces (60J27)

Cites Work

• The ECME algorithm: A simple extension of EM and ECM with faster monotone convergence
• Title not available (Why is that?)
• Numerical analysis for statisticians
• Numerical recipes. The art of scientific computing.
• Stochastic epidemic models and their statistical analysis
• Mathematical biology. Vol. 1: An introduction.
• Numerical Inversion of Laplace Transforms of Probability Distributions
• Continued fractions with applications
• Numerical inversion of probability generating functions
• Transition probabilities for general birth-death processes with applications in ecology, genetics, and evolution
• Title not available (Why is that?)
• Generator estimation of Markov jump processes
• Statistical Inference for Discretely Observed Markov Jump Processes
• Ancestral processes with selection
• Stochastic population processes. Analysis, approximation, simulations.
• The Fourier-series method for inverting transforms of probability distributions
• Direct Calculation of the Information Matrix via the EM Algorithm
• Maximum likelihood estimation in the birth-and-death process
• Computing Laplace Transforms for Numerical Inversion Via Continued Fractions
• Development of non-linear transformations for improving convergence of sequences
• Problems of Statistical Inference for Birth and Death Queuing Models
• Title not available (Why is that?)
• Estimation in the birth process
• A survey of truncation error analysis for Padé and continued fraction approximants
• Exact transient solution of a state-dependent birth-death process
• Some Properties of Continued Fractions with Applications in Markov Processes
• The formal theory of birth-and-death processes, lattice path combinatorics and continued fractions
• Application of Stieltjes theory for S-fractions to birth and death processes
• Excursions of birth and death processes, orthogonal polynomials, and continued fractions
• Numerical Evaluation of Continued Fractions
• Density estimation in high and ultra high dimensions, regularization, and the \(L_1\) asymptotics
• Numerical Continued Fractions
• Simulation from endpoint-conditioned, continuous-time Markov chains on a finite state space, with applications to molecular evolution
• THE BEHAVIOUR OF AN ESTIMATOR FOR A SIMPLE BIRTH AND DEATH PROCESS
• Estimating change rates of genetic markers using serial samples: applications to the transposon IS6110 in Mycobacterium tuberculosis
• Continued Fraction Analysis of the Duration of an Excursion in an M/M/∞ System
• Graphics processing units and high-dimensional optimization
• Counting labeled transitions in continuous-time Markov models of evolution
• Miscellanea. Information matrix computation from conditional information via normal approximation
• Bayesian inference for linear growth birth and death processes.
• Fitting birth-death processes to panel data with applications to bacterial DNA fingerprinting
• Summary Statistics for Endpoint-Conditioned Continuous-Time Markov Chains

Cited In (22)

• On a reward rate estimation for the finite irreducible continuous-time Markov chain
• Parameter estimation for multivariate population processes: a saddlepoint approach
• Birth/birth-death processes and their computable transition probabilities with biological applications
• Transition probabilities for general birth-death processes with applications in ecology, genetics, and evolution
• Title not available (Why is that?)
• Computational methods for birth-death processes
• Sex, lies and self-reported counts: Bayesian mixture models for heaping in longitudinal count data via birth-death processes
• A three-term recurrence relation for accurate evaluation of transition probabilities of the simple birth-and-death process
• The probability distribution of the ancestral population size conditioned on the reconstructed phylogenetic tree with occurrence data
• Gain-loss-duplication models for copy number evolution on a phylogeny: exact algorithms for computing the likelihood and its gradient
• INAR approximation of bivariate linear birth and death process
• A numerical approach for evaluating the time-dependent distribution of a quasi birth-death process
• Coupling bounds for approximating birth-death processes by truncation
• Some mechanisms leading to underdispersion: Old and new proposals
• Probabilistic predictions of SIS epidemics on networks based on population-level observations
• Title not available (Why is that?)
• Parameter estimation for discretely observed linear birth‐and‐death processes
• Fitting birth-death processes to panel data with applications to bacterial DNA fingerprinting
• Inferring density-dependent population dynamics mechanisms through rate disambiguation for logistic birth-death processes
• Likelihood-based inference for discretely observed birth-death-shift processes, with applications to evolution of mobile genetic elements
• The linear birth-death process: an inferential retrospective
• Towards a statistical physics of dating apps

This page was built for publication: Estimation for general birth-death processes

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4975412)