DOCSLIB.ORG

Uncovering the Temporal Dynamics of Diffusion Networks

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Uncovering the Temporal Dynamics of Diffusion Networks Uncovering the Temporal Dynamics of Diffusion Networks Manuel Gomez-Rodriguez1;2 [email protected] David Balduzzi1 [email protected] Bernhard Scholkopf¨ 1 [email protected] 1MPI for Intelligent Systems and 2Stanford University Abstract Observing a diffusion process often reduces to noting when Time plays an essential role in the diffusion of in- nodes (people, blogs, etc.) reproduce a piece of informa- formation, influence and disease over networks. tion, get infected by a virus, or buy a product. Epidemiolo- In many cases we only observe when a node gists can observe when a person becomes ill but they can- copies information, makes a decision or becomes not tell who infected her or how many exposures and how infected – but the connectivity, transmission rates much time was necessary for the infection to take hold. In between nodes and transmission sources are un- information propagation, we observe when a blog mentions known. Inferring the underlying dynamics is a piece of information. However if, as is often the case, of outstanding interest since it enables forecast- the blogger does not link to her source, we do not know ing, influencing and retarding infections, broadly where she acquired the information or how long it took construed. To this end, we model diffusion pro- her to post it. Finally, viral marketers can track when cus- cesses as discrete networks of continuous tempo- tomers buy products or subscribe to services, but typically ral processes occurring at different rates. Given cannot observe who influenced customers’ decisions, how cascade data – observed infection times of nodes long they took to make up their minds, or when they passed – we infer the edges of the global diffusion net- recommendations on to other customers. In all these sce- work and estimate the transmission rates of each narios, we observe where and when but not how or why edge that best explain the observed data. The op- information (be it in the form of a virus, a meme, or a timization problem is convex. The model nat- decision) propagates through a population of individuals. urally (without heuristics) imposes sparse solu- The mechanism underlying the process is hidden. How- tions and requires no parameter tuning. The ever, the mechanism is of outstanding interest in all three problem decouples into a collection of indepen- cases, since understanding diffusion is necessary for stop- dent smaller problems, thus scaling easily to net- ping infections, predicting meme propagation, or maximiz- works on the order of hundreds of thousands of ing sales of a product. nodes. Experiments on real and synthetic data This article presents a method for inferring the mechanisms show that our algorithm both recovers the edges underlying diffusion processes based on observed infec- of diffusion networks and accurately estimates tions. To achieve this aim, we construct a model incor- their transmission rates from cascade data. porating some basic assumptions about the spatiotemporal structures that generate diffusion processes. The assump- tions are as follows. First, diffusion processes occur over 1. Introduction static (fixed) but unknown networks (directed graphs). Sec- ond, infections are binary, i.e., a node is either infected or Diffusion and propagation processes have received increas- it is not; we do not model partial infections or the partial ing attention in a broad range of domains: information propagation of information. Third, infections along edges propagation (Adar & Adamic, 2005; Gomez-Rodriguez of the network occur independently of each other. Fourth, et al., 2010; Meyers & Leskovec, 2010), social net- an infection can occur at different times: the likelihood of works (Kempe et al., 2003; Lappas et al., 2010), viral mar- node a infecting node b at time t is modeled via a proba- keting (Watts & Dodds, 2007) and epidemiology (Wallinga bility density function depending on a, b and t. Finally, we & Teunis, 2004). observe all infections occurring in the network during the Appearing in Proceedings of the 28 th International Conference recorded time window. Our aim is to infer the connectiv- on Machine Learning, Bellevue, WA, USA, 2011. Copyright 2011 ity of the network and the likelihood of infections across by the author(s)/owner(s). its edges after observing the times at which nodes in the Uncovering the Temporal Dynamics of Diffusion Networks network become infected. section we formulate our model, starting from the data it is designed for, and concluding with a precise statement of In more detail, we formulate a generative probabilistic the network inference problem. model of diffusion that aims to describe realistically how infections occur over time in a static network. Finding Data. Observations are recorded on a fixed population of the optimal network and transmission rates maximizing the N nodes and consist of a set C of cascades ft1;:::; tjCjg. likelihood of an observed set of infection cascades reduces Each cascade tc is a record of observed infection times to solving a convex program. The convex problem decou- within the population during a time interval of length T c. ples into many smaller problems, allowing for natural par- c c c A cascade is an N-dimensional vector t := (t1; : : : ; tN ) allelization so that our algorithm scales to networks with c c recording when nodes are infected, tk 2 [0;T ] [ f1g. hundreds of thousands of nodes. We show the effectiveness Symbol 1 labels nodes that are not infected during ob- of our method by reconstructing the connectivity and con- servation window [0;T c] – it does not imply that nodes tinuous temporal dynamics of synthetic and real networks are never infected. The ‘clock’ is reset to 0 at the start using cascade data. of each cascade. Lengthening the observation window T c increases the number of observed infections within a cas- Related work. The work most closely related to cade c and results in a more representative sample of the ours (Gomez-Rodriguez et al., 2010; Meyers & Leskovec, underlying dynamics. However, these advantages must be 2010) also uses a generative probabilistic model for infer- weighed against the cost of observing for longer periods. ring diffusion networks. Gomez-Rodriguez et al. (2010) For simplicity we assume T c = T for all cascades; the (NETINF) infers network connectivity using submodular results generalize trivially. optimization and Meyers & Leskovec (2010) (CONNIE) infer not only the connectivity but also a prior probabil- The time-stamps assigned to nodes by a cascade induce the ity of infection for every edge using a convex program and structure of a directed acyclic graph (DAG) on the network some heuristics. However, both papers force the transmis- (which is not acyclic in general) by defining node i is a par- sion rate between all nodes to be fixed – and not inferred. ent of j if ti < tj. Thus, it is meaningful to refer to parents In contrast, our model allows transmission at different rates and children within a cascade, but not on the network. The across different edges so that we can infer temporally het- DAG structure dramatically simplifies the computational erogeneous interactions within a network, as found in real- complexity of the inference problem. Also, since the un- world examples. Thus, we can now infer the temporal dy- derlying network is inferred from many cascades (each of namics of the underlying network. which imposes its own DAG structure), the inferred net- work is typically not a DAG. The main innovation of this paper is to model diffusion as a spatially discrete network of continuous, conditionally in- Pairwise transmission likelihood. The first step in model- dependent temporal processes occurring at different rates. ing diffusion dynamics is to consider pairwise interactions. Infection transmission depends on the complex intricacies We assume that infections can occur at different rates over e.g. of the underlying mechanisms ( , a person’s susceptibil- different edges of a network, and aim to infer the transmis- ity to viral infections depends on weather, diet, age, stress sion rates between pairs of nodes in the network. levels, prior exposures to similar pathogens and so on). We avoid modeling the mechanisms underlying individual in- Define f(tijtj; αj;i) as the conditional likelihood of trans- fections, and instead develop a data-driven approach, suit- mission between a node j and node i. The transmission able for large-scale analyses, that infers the diffusion pro- likelihood depends on the infection times (tj; ti) and a pair- cess using only the visible spatiotemporal traces (cascades) wise transmission rate αj;i. A node cannot be infected by a it generates. We therefore model diffusion using only time- node infected later in time. In other words, a node j that has dependent pairwise transmission likelihood between pairs been infected at a time tj may infect a node i at a time ti of nodes, transmission rates and infection times, but not only if tj < ti. Although in some scenarios it may be possi- prior probabilities of infection that depend on unknown ble to estimate a non-parametric likelihood empirically, for external factors. To the best of our knowledge, continu- simplicity we consider three well-known parametric mod- ous temporal dynamics of diffusion networks has not been els: exponential, power-law and Rayleigh (see Table 1). modeled or inferred in previous work. We believe this is a Transmission rates are denoted as αj;i ≥ 0 and δ is the key point for understanding diffusion processes. minimum allowed time difference in the power-law to have a bounded likelihood. As αj;i ! 0 the likelihood of in- 2. Problem formulation fection tends to zero and the expected transmission time becomes arbitrarily long. Without loss of generality, we This paper develops a method for inferring the spatiotem- consider δ = 1 in the power-law model from now on.
Load more

Recommended publications
  • A Stochastic Processes and Martingales
    A Stochastic Processes and Martingales A.1 Stochastic Processes Let I be either IINorIR+.Astochastic process on I with state space E is a family of E-valued random variables X = {Xt : t ∈ I}. We only consider examples where E is a Polish space. Suppose for the moment that I =IR+. A stochastic process is called cadlag if its paths t → Xt are right-continuous (a.s.) and its left limits exist at all points. In this book we assume that every stochastic process is cadlag. We say a process is continuous if its paths are continuous. The above conditions are meant to hold with probability 1 and not to hold pathwise. A.2 Filtration and Stopping Times The information available at time t is expressed by a σ-subalgebra Ft ⊂F.An {F ∈ } increasing family of σ-algebras t : t I is called a filtration.IfI =IR+, F F F we call a filtration right-continuous if t+ := s>t s = t. If not stated otherwise, we assume that all filtrations in this book are right-continuous. In many books it is also assumed that the filtration is complete, i.e., F0 contains all IIP-null sets. We do not assume this here because we want to be able to change the measure in Chapter 4. Because the changed measure and IIP will be singular, it would not be possible to extend the new measure to the whole σ-algebra F. A stochastic process X is called Ft-adapted if Xt is Ft-measurable for all t. If it is clear which filtration is used, we just call the process adapted.The {F X } natural filtration t is the smallest right-continuous filtration such that X is adapted.
    [Show full text]
  • 6 Mar 2019 Strict Local Martingales and the Khasminskii Test for Explosions
    Strict Local Martingales and the Khasminskii Test for Explosions Aditi Dandapani∗† and Philip Protter‡§ March 7, 2019 Abstract We exhibit sufficient conditions such that components of a multidimen- sional SDE giving rise to a local martingale M are strict local martingales or martingales. We assume that the equations have diffusion coefficients of the form σ(Mt, vt), with vt being a stochastic volatility term. We dedicate this paper to the memory of Larry Shepp 1 Introduction In 2007 P.L. Lions and M. Musiela (2007) gave a sufficient condition for when a unique weak solution of a one dimensional stochastic volatility stochastic differen- tial equation is a strict local martingale. Lions and Musiela also gave a sufficient condition for it to be a martingale. Similar results were obtained by L. Andersen and V. Piterbarg (2007). The techniques depend at one key stage on Feller’s test for explosions, and as such are uniquely applicable to the one dimensional case. Earlier, in 2002, F. Delbaen and H. Shirakawa (2002) gave a seminal result giving arXiv:1903.02383v1 [q-fin.MF] 6 Mar 2019 necessary and sufficient conditions for a solution of a one dimensional stochastic differential equation (without stochastic volatility) to be a strict local martingale or not. This was later extended by A. Mijatovic and M. Urusov (2012) and then by ∗Applied Mathematics Department, Columbia University, New York, NY 10027; email: [email protected]; Currently at Ecole Polytechnique, Palaiseau, France. †Supported in part by NSF grant DMS-1308483 ‡Statistics Department, Columbia University, New York, NY 10027; email: [email protected].
    [Show full text]
  • Geometric Brownian Motion with Affine Drift and Its Time-Integral
    Geometric Brownian motion with affine drift and its time-integral ⋆ a b, c Runhuan Feng , Pingping Jiang ∗, Hans Volkmer aDepartment of Mathematics, University of Illinois at Urbana-Champaign, Illinois, USA bSchool of Management and Economics,The Chinese University of Hong Kong, Shenzhen, Shenzhen, Guangdong, 518172, P.R. China School of Management, University of Science and Technology of China, Hefei, Anhui, 230026, P.R.China cDepartment of Mathematical Sciences, University of Wisconsin–Milwaukee, Wisconsin, USA Abstract The joint distribution of a geometric Brownian motion and its time-integral was derived in a seminal paper by Yor (1992) using Lamperti’s transformation, leading to explicit solutions in terms of modified Bessel functions. In this paper, we revisit this classic result using the simple Laplace transform approach in connection to the Heun differential equation. We extend the methodology to the geometric Brownian motion with affine drift and show that the joint distribution of this process and its time-integral can be determined by a doubly-confluent Heun equation. Furthermore, the joint Laplace transform of the process and its time-integral is derived from the asymptotics of the solutions. In addition, we provide an application by using the results for the asymptotics of the double-confluent Heun equation in pricing Asian options. Numerical results show the accuracy and efficiency of this new method. Keywords: Doubly-confluent Heun equation, geometric Brownian motion with affine drift, Lamperti’s transformation, asymptotics, boundary value problem. ⋆R. Feng is supported by an endowment from the State Farm Companies Foundation. P. Jiang arXiv:2012.09661v1 [q-fin.MF] 17 Dec 2020 is supported by the Postdoctoral Science Foundation of China (No.2020M671853) and the National Natural Science Foundation of China (No.
    [Show full text]
  • Jump-Diffusion Models for Asset Pricing in Financial Engineering
    J.R. Birge and V. Linetsky (Eds.), Handbooks in OR & MS, Vol. 15 Copyright © 2008 Elsevier B.V. All rights reserved DOI: 10.1016/S0927-0507(07)15002-7 Chapter 2 Jump-Diffusion Models for Asset Pricing in Financial Engineering S.G. Kou Department of Industrial Engineering and Operations Research, Columbia University E-mail: [email protected] Abstract In this survey we shall focus on the following issues related to jump-diffusion mod- els for asset pricing in financial engineering. (1) The controversy over tailweight of distributions. (2) Identifying a risk-neutral pricing measure by using the rational ex- pectations equilibrium. (3) Using Laplace transforms to pricing options, including European call/put options, path-dependent options, such as barrier and lookback op- tions. (4) Difficulties associated with the partial integro-differential equations related to barrier-crossing problems. (5) Analytical approximations for finite-horizon Amer- ican options with jump risk. (6) Multivariate jump-diffusion models. 1Introduction There is a large literature on jump-diffusion models in finance, including several excellent books, e.g. the books by Cont and Tankov (2004), Kijima (2002). So a natural question is why another survey article is needed. What we attempt to achieve in this survey chapter is to emphasis some points that have not been well addressed in previous surveys. More precisely we shall focus on the following issues. (1) The controversy over tailweight of distributions. An empirical motiva- tion for using jump-diffusion models comes from the fact that asset return distributions tend to have heavier tails than those of normal dis- tribution. However, it is not clear how heavy the tail distributions are, as some people favor power-type distributions, others exponential-type distributions.
    [Show full text]
  • 1 the Ito Integral
    Derivative Securities, Fall 2010 Mathematics in Finance Program Courant Institute of Mathematical Sciences, NYU Jonathan Goodman http://www.math.nyu.edu/faculty/goodman/teaching/DerivSec10/resources.html Week 6 1 The Ito integral The Black Scholes reasoning asks us to apply calculus, stochastic calculus, to expressions involving differentials of Brownian motion and other diffusion pro- cesses. To do this, we define the Ito integral with respect to Brownian motion and a general diffusion. Things discussed briefly and vaguely here are discussed in more detail and at greater length in Stochastic calculus. Suppose Xt is a diffusion process that satisfies (18) and (19) from week 5. Suppose that ft is another stochastic process that is adapted to the same filtration F. The Ito integral is Z t Yt = fs dXs : (1) 0 You could call this the \indefinite integral" because we care how the answer depends on t. In particular, the Ito integral is one of the ways to construct a new stochastic process, Yt, from old ones ft and Xt. It is not possible to define (1) unless ft is adapted. If ft is allowed to depend 0 on future values Xt0 (t > t), then the integral may not make sense or it may not have the properties we expect. The essential technical idea in the definition of (1) is that dXs is in the future of s. If we think that dXs = Xs+ds − Xs, then we must take ds > 0 (but infinitely close to zero). This implies that if Xt is a martingale, then Yt also is a martingale.
    [Show full text]
  • A Stochastic Lomax Diffusion Process: Statistical Inference and Application
    mathematics Article A Stochastic Lomax Diffusion Process: Statistical Inference and Application Ahmed Nafidi 1,†, Ilyasse Makroz 1,*,† and Ramón Gutiérrez Sánchez 2,† 1 Department of Mathematics and Informatics, National School of Applied Sciences, Hassan First University of Settat, LAMSAD, Avenue de l’université, Berrechid BP 280, Morocco; ahmed.nafi[email protected] 2 Department of Statistics and Operational Research, Facultad de Ciencias, Campus de Fuentenueva, University of Granada, 18071 Granada, Spain; [email protected] * Correspondence: [email protected] † These authors contributed equally to this work. Abstract: In this paper, we discuss a new stochastic diffusion process in which the trend function is proportional to the Lomax density function. This distribution arises naturally in the studies of the frequency of extremely rare events. We first consider the probabilistic characteristics of the proposed model, including its analytic expression as the unique solution to a stochastic differential equation, the transition probability density function together with the conditional and unconditional trend functions. Then, we present a method to address the problem of parameter estimation using maximum likelihood with discrete sampling. This estimation requires the solution of a non-linear equation, which is achieved via the simulated annealing method. Finally, we apply the proposed model to a real-world example concerning adolescent fertility rate in Morocco. Keywords: stochastic differential equation; lomax distribution; trend functions; statistical inference; simulated annealing; adolescent fertility rate Citation: Nafidi, A.; Makroz, I.; 1. Introduction Gutiérrez-Sánchez, R. A Stochastic Stochastic diffusion models are used to analyze the evolution of phenomena in multi- Lomax Diffusion Process: Statistical ple fields of science, including biology, finance, energy consumption and physics.
    [Show full text]
  • A Guide to Brownian Motion and Related Stochastic Processes
    Vol. 0 (0000) A guide to Brownian motion and related stochastic processes Jim Pitman and Marc Yor Dept. Statistics, University of California, 367 Evans Hall # 3860, Berkeley, CA 94720-3860, USA e-mail: [email protected] Abstract: This is a guide to the mathematical theory of Brownian mo- tion and related stochastic processes, with indications of how this theory is related to other branches of mathematics, most notably the classical the- ory of partial differential equations associated with the Laplace and heat operators, and various generalizations thereof. As a typical reader, we have in mind a student, familiar with the basic concepts of probability based on measure theory, at the level of the graduate texts of Billingsley [43] and Durrett [106], and who wants a broader perspective on the theory of Brow- nian motion and related stochastic processes than can be found in these texts. Keywords and phrases: Markov process, random walk, martingale, Gaus- sian process, L´evy process, diffusion. AMS 2000 subject classifications: Primary 60J65. Contents 1 Introduction................................. 3 1.1 History ................................ 3 1.2 Definitions............................... 4 2 BM as a limit of random walks . 5 3 BMasaGaussianprocess......................... 7 3.1 Elementarytransformations . 8 3.2 Quadratic variation . 8 3.3 Paley-Wiener integrals . 8 3.4 Brownianbridges........................... 10 3.5 FinestructureofBrownianpaths . 10 arXiv:1802.09679v1 [math.PR] 27 Feb 2018 3.6 Generalizations . 10 3.6.1 FractionalBM ........................ 10 3.6.2 L´evy’s BM . 11 3.6.3 Browniansheets ....................... 11 3.7 References............................... 11 4 BMasaMarkovprocess.......................... 12 4.1 Markovprocessesandtheirsemigroups. 12 4.2 ThestrongMarkovproperty. 14 4.3 Generators .............................
    [Show full text]
  • ENLARGEMENT of FILTRATION and the STRICT LOCAL MARTINGALE PROPERTY in STOCHASTIC DIFFERENTIAL EQUATIONS Aditi Dandapani COLUMBIA
    ENLARGEMENT OF FILTRATION AND THE STRICT LOCAL MARTINGALE PROPERTY IN STOCHASTIC DIFFERENTIAL EQUATIONS Aditi Dandapani Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Graduate School of Arts and Sciences COLUMBIA UNIVERSITY 2016 © 2016 Aditi Dandapani All Rights Reserved Abstract ENLARGEMENT OF FILTRATION AND THE STRICT LOCAL MARTINGALE PROPERTY IN STOCHASTIC DIFFERENTIAL EQUATIONS Aditi Dandapani In this thesis, we study the strict local martingale property of solutions of various types of stochastic differential equations and the effect of an initial expansion of the filtration on this property. For the models we consider, we either use existing criteria or, in the case where the stochastic differential equation has jumps, develop new criteria that can can detect the presence of the strict local martingale property. We develop deterministic sufficient conditions on the drift and diffusion coefficient of the stochastic process such that an enlargement by initial expansion of the filtration can produce a strict local martingale from a true martingale. We also develop a way of characterizing the martingale property in stochastic volatility models where the local martingale has a general diffusion coefficient, of the form µ(St; vt);where the function µ(x1; x2) is locally Lipschitz in x: Contents Acknowledgements ii Introduction 1 Chapter 1 5 0.1 The One-Dimensional Case . .5 0.2 Stochastic Volatility Models . 13 0.3 Expansion of the Filtration by Initial Expansions . 31 Chapter 2 35 0.4 The Model of Lions and Musiela . 35 0.5 The Case of Jump Discontinuities . 62 0.6 The Model of Mijatovic and Urusov .
    [Show full text]
  • Poisson Processes and Jump Diffusion Model
    Mälardalen University MT 1410 Analytical Finance I Teacher: Jan Röman, IMA Seminar 2005-10-05 Poisson Processes and Jump Diffusion Model Authors: Ying Ni, Cecilia Isaksson 13/11/2005 Abstract Besides the Wiener process (Brownian motion), there’s an another stochastic process that is very useful in finance, the Poisson process. In a Jump diffusion model, the asset price has jumps superimposed upon the familiar geometric Brownian motion. The occurring of those jumps are modelled using a Poisson process. This paper introduces the definition and properties of Poisson process and thereafter the Jump diffusion process which consists two stochastic components, the “white noise” created by Wiener process and the jumps generated by Poisson processes. MT1410 Seminar Group: Cecilia Isaksson; Ying Ni - 1 - 13/11/2005 Table of contents 1. Introduction …..………………………………………………………………….3 2. Poisson Processes ……………………………………………………………….4 2.1 Lévy Processes, Wiener Processes & Poisson Processes ……………………….4 2.2 The Markov Property of Poisson Processes ……………………………………..6 2.2 Applications of Poisson Processes in Finance……………………………………..8 3. The Jump Diffusion Model …………………………………………………….9 3.1 The model…………………………………………………………………………..9 3.2 Practical problem………………………………………………………………….10 4. Conclusion ………………………………………………………………………11 5. References ………………………………………………………………………..12 MT1410 Seminar Group: Cecilia Isaksson; Ying Ni - 2 - 13/11/2005 1. Introduction In option pricing theory, the geometric Brownian Motion is the most frequently used model for the price dynamic of an underlying asset. However, it is argued that this model fails to capture properly the unexpected price changes, called jumps. Price jumps are important because they do exist in the market. A more realistic model should therefore also take jumps into account. Price jumps are in general infrequent events and therefore the number of those jumps can be modelled by a Poisson processes.
    [Show full text]
  • Martingale Problems and Stochastic Equations
    Martingale problems and stochastic equations • Characterizing stochastic processes by their martingale properties • Markov processes and their generators • Forward equations • Other types of martingale problems • Change of measure • Martingale problems for conditional distributions • Stochastic equations for Markov processes • Filtering • Time change equations • Bits and pieces • Supplemental material • Exercises • References •First •Prev •Next •Go To •Go Back •Full Screen •Close •Quit 1 Supplemental material • Background: Basics of stochastic processes • Stochastic integrals for Poisson random measures • Technical lemmas •First •Prev •Next •Go To •Go Back •Full Screen •Close •Quit 2 1. Characterizing stochastic processes by their martingale prop- erties • Levy’s´ characterization of Brownian motion • Definition of a counting process • Poisson processes • Martingale properties of the Poisson process • Strong Markov property for the Poisson process • Intensities • Counting processes as time changes of Poisson processes • Martingale characterizations of a counting process • Multivariate counting processes Jacod(1974/75),Anderson and Kurtz(2011, 2015) •First •Prev •Next •Go To •Go Back •Full Screen •Close •Quit 3 Brownian motion A Brownian motion is a continuous process with independent, Gaussian increments. A Brownian motion W is standard if the increments W (t + r) − W (t), t; r ≥ 0, have mean zero and variance r. W is a martingale: W W E[W (t + r)jFt ] = E[W (t + r) − W (t)jFt ] + W (t) = W (t); W Ft = σ(W (s): s ≤ t). W has quadratic variation t, that is X 2 [W ]t = lim (W (t ^ ti+1) − W (t ^ ti)) = t sup jti+1−tij=0 i in probability, or more precisely, X 2 lim sup j (W (t ^ ti+1) − W (t ^ ti)) − tj = 0 sup jti+1−tij=0 t≤T i in probability for each T > 0.
    [Show full text]
  • On a Class of Historical Superprocesses in the Dynkin Sense
    View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by Saitama University Cyber Repository of Academic Resources J. Saitama Univ. Fac. Educ., 62(1)：249-266 （2013） On A Class of Historical Superprocesses in The Dynkin Sense DOKU,^ Isamu Faculty of Education, Saitama University Abstract We study a class of historical superprocesses in the Dynkin sense. This class is closely related to a group of superprocesses, namely, measure-valued branching Markov processes, which are associated with stable random measure. Then we prove the existence theorem for the class of historical superprocesses in Dynkin's sense. Key Words : Dynkin's historical superprocess, measure-valued Markov process, superprocess, stable random measure, existence. 1. Introduction The purpose of this paper is to investigate a class of historical superprocesses in Dynkin's sense. This class is closely related to another class of superprocesses, namely, measure-valued branching Markov processes, which are associated with stable random measure. Then it is proven that the class of historical superprocesses in Dynkin's sense exists under some suitable conditions. Now we shall begin with introducing a fundamental relationship between superprocess and its associated historical superprocess. 1.1 Superprocess with Branching Rate Functional In this subsection we shall introduce the so-called superprocess with branching rate functional, which forms a general class of measure-valued branching Markov processes with diffusion as a underlying spatial motion. WeR write the integral of measurable function f with respect to measure µ as hµ, fi = fdµ. For simplicity, d d MF = MF (R ) is the space of finite measures on R .
    [Show full text]
  • Week 8 Diffusion Processes, Part 2
    Week 8 Diffusion processes, part 2 Jonathan Goodman October 28, 2013 1 Integration and Ito's lemma for dXt sec:il Outline of this section: 1. Ways to define new processes using old ones: R t (a) An Ito integral with respect to an old process, Yt = 0 asdXs (b) A function of an old process, Yt = f(Xt; t) 2. Figure out stuff about the new processes (a) Ito's lemma 3. It is more useful when it is more general (a) General Ito process, not just diffusions (b) Multi-component processes 1.1 Technical overview A diffusion is a process that satisfies an SDE, which makes it a Markov process. The operations listed above produce processes that have Ito differentials but are not Markov. We call Xt an Ito process if (in these formulas, ∆t > 0 and ∆X = Xt+∆t − Xt): E[ ∆X j Ft] = at∆t + o(∆t) (1) eq:im h 2 i E (∆X) j Ft = µt∆t + o(∆t) (one component) (2) eq:iv h t i E (∆X) (∆X) j Ft = µt∆t + o(∆t) (multi-component) (3) eq:ivm h 4 i 2 E j∆Xj j Ft ≤ C∆t (4) eq:m4 The little oh notation A = o(∆t) means that A is smaller than ∆t. More precisely, A=∆t ! 0 as ∆t ! 0. It is usually correct to think O(∆t2) whenever 1 you see o(∆t). That's what it would be if it were a Taylor series expansion. But it is convenient to make do with an error estimate that is weaker than O(∆t2).
    [Show full text]