Track Properities of Normal Chain

Similar documents
GENERATING CERTAIN QUINTIC IRREDUCIBLE POLYNOMIALS OVER FINITE FIELDS. Youngwoo Ahn and Kitae Kim

Existence and Uniqueness Results for Random Impulsive Integro-Differential Equation

EP2200 Queuing theory and teletraffic systems. 3rd lecture Markov chains Birth-death process - Poisson process. Viktoria Fodor KTH EES

John Geweke a and Gianni Amisano b a Departments of Economics and Statistics, University of Iowa, USA b European Central Bank, Frankfurt, Germany

On One Analytic Method of. Constructing Program Controls

Comparison of Differences between Power Means 1

V.Abramov - FURTHER ANALYSIS OF CONFIDENCE INTERVALS FOR LARGE CLIENT/SERVER COMPUTER NETWORKS

Lecture 2 M/G/1 queues. M/G/1-queue

Relative controllability of nonlinear systems with delays in control

SOME NOISELESS CODING THEOREMS OF INACCURACY MEASURE OF ORDER α AND TYPE β

CH.3. COMPATIBILITY EQUATIONS. Continuum Mechanics Course (MMC) - ETSECCPB - UPC

Testing a new idea to solve the P = NP problem with mathematical induction

Notes on the stability of dynamic systems and the use of Eigen Values.

CS286.2 Lecture 14: Quantum de Finetti Theorems II

Ordinary Differential Equations in Neuroscience with Matlab examples. Aim 1- Gain understanding of how to set up and solve ODE s

Appendix H: Rarefaction and extrapolation of Hill numbers for incidence data

e-journal Reliability: Theory& Applications No 2 (Vol.2) Vyacheslav Abramov

Part II CONTINUOUS TIME STOCHASTIC PROCESSES

Mechanics Physics 151

Solution in semi infinite diffusion couples (error function analysis)

Approximate Analytic Solution of (2+1) - Dimensional Zakharov-Kuznetsov(Zk) Equations Using Homotopy

Comb Filters. Comb Filters

A New Generalized Gronwall-Bellman Type Inequality

CS434a/541a: Pattern Recognition Prof. Olga Veksler. Lecture 4

5th International Conference on Advanced Design and Manufacturing Engineering (ICADME 2015)

Fuzzy Set Theory in Modeling Uncertainty Data. via Interpolation Rational Bezier Surface Function

Let s treat the problem of the response of a system to an applied external force. Again,

Volatility Interpolation

Tight results for Next Fit and Worst Fit with resource augmentation

HEAT CONDUCTION PROBLEM IN A TWO-LAYERED HOLLOW CYLINDER BY USING THE GREEN S FUNCTION METHOD

Case Study of Markov Chains Ray-Knight Compactification

In the complete model, these slopes are ANALYSIS OF VARIANCE FOR THE COMPLETE TWO-WAY MODEL. (! i+1 -! i ) + [(!") i+1,q - [(!

( ) () we define the interaction representation by the unitary transformation () = ()

Name of the Student:

( t) Outline of program: BGC1: Survival and event history analysis Oslo, March-May Recapitulation. The additive regression model

UNIVERSITAT AUTÒNOMA DE BARCELONA MARCH 2017 EXAMINATION

GORDON AND NEWELL QUEUEING NETWORKS AND COPULAS

Online Supplement for Dynamic Multi-Technology. Production-Inventory Problem with Emissions Trading

Mechanics Physics 151

ON THE WEAK LIMITS OF SMOOTH MAPS FOR THE DIRICHLET ENERGY BETWEEN MANIFOLDS

Mechanics Physics 151

How about the more general "linear" scalar functions of scalars (i.e., a 1st degree polynomial of the following form with a constant term )?

Survival Analysis and Reliability. A Note on the Mean Residual Life Function of a Parallel System

@FMI c Kyung Moon Sa Co.

Dynamic Team Decision Theory. EECS 558 Project Shrutivandana Sharma and David Shuman December 10, 2005

FI 3103 Quantum Physics

Lecture VI Regression

Graduate Macroeconomics 2 Problem set 5. - Solutions

Bayesian Inference of the GARCH model with Rational Errors

Should Exact Index Numbers have Standard Errors? Theory and Application to Asian Growth

Existence of Time Periodic Solutions for the Ginzburg-Landau Equations. model of superconductivity

The Properties of Probability of Normal Chain

Cubic Bezier Homotopy Function for Solving Exponential Equations

Sampling Procedure of the Sum of two Binary Markov Process Realizations

P R = P 0. The system is shown on the next figure:

Math 128b Project. Jude Yuen

An introduction to Support Vector Machine

( ) [ ] MAP Decision Rule

THE PREDICTION OF COMPETITIVE ENVIRONMENT IN BUSINESS

Advanced Machine Learning & Perception

CHAPTER 10: LINEAR DISCRIMINATION

On the numerical treatment ofthenonlinear partial differentialequation of fractional order

Chapters 2 Kinematics. Position, Distance, Displacement

COMPUTER SCIENCE 349A SAMPLE EXAM QUESTIONS WITH SOLUTIONS PARTS 1, 2

Method of upper lower solutions for nonlinear system of fractional differential equations and applications

This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and

Theoretical Analysis of Biogeography Based Optimization Aijun ZHU1,2,3 a, Cong HU1,3, Chuanpei XU1,3, Zhi Li1,3

NATIONAL UNIVERSITY OF SINGAPORE PC5202 ADVANCED STATISTICAL MECHANICS. (Semester II: AY ) Time Allowed: 2 Hours

KEY. Math 334 Midterm III Winter 2008 section 002 Instructor: Scott Glasgow

Department of Economics University of Toronto

SELFSIMILAR PROCESSES WITH STATIONARY INCREMENTS IN THE SECOND WIENER CHAOS

Introduction ( Week 1-2) Course introduction A brief introduction to molecular biology A brief introduction to sequence comparison Part I: Algorithms

arxiv: v1 [math.pr] 6 Mar 2019

Political Economy of Institutions and Development: Problem Set 2 Due Date: Thursday, March 15, 2019.

Epistemic Game Theory: Online Appendix

Tools for Analysis of Accelerated Life and Degradation Test Data

Lecture 6: Learning for Control (Generalised Linear Regression)

Uniform Topology on Types and Strategic Convergence

Lecture 18: The Laplace Transform (See Sections and 14.7 in Boas)

Econ107 Applied Econometrics Topic 5: Specification: Choosing Independent Variables (Studenmund, Chapter 6)

[ ] 2. [ ]3 + (Δx i + Δx i 1 ) / 2. Δx i-1 Δx i Δx i+1. TPG4160 Reservoir Simulation 2018 Lecture note 3. page 1 of 5

Chapter 6: AC Circuits

arxiv: v1 [cs.sy] 2 Sep 2014

A Deza Frankl type theorem for set partitions

Lecture 1 Overview. course mechanics. outline & topics. what is a linear dynamical system? why study linear systems? some examples

2/20/2013. EE 101 Midterm 2 Review

Scattering at an Interface: Oblique Incidence

Transient Response in Electric Circuits

Chapter Lagrangian Interpolation

Consensus of Multi-agent Systems Under Switching Agent Dynamics and Jumping Network Topologies

Density Matrix Description of NMR BCMB/CHEM 8190

Implementation of Quantized State Systems in MATLAB/Simulink

. The geometric multiplicity is dim[ker( λi. A )], i.e. the number of linearly independent eigenvectors associated with this eigenvalue.

H = d d q 1 d d q N d d p 1 d d p N exp

I-POLYA PROCESS AND APPLICATIONS Leda D. Minkova

10. A.C CIRCUITS. Theoretically current grows to maximum value after infinite time. But practically it grows to maximum after 5τ. Decay of current :

The safety stock and inventory cost paradox in a stochastic lead time setting

Lecture 11 SVM cont

Endpoint Strichartz estimates

A DECOMPOSITION METHOD FOR SOLVING DIFFUSION EQUATIONS VIA LOCAL FRACTIONAL TIME DERIVATIVE

Transcription:

In. J. Conemp. Mah. Scences, Vol. 8, 213, no. 4, 163-171 HIKARI Ld, www.m-har.com rac Propes of Normal Chan L Chen School of Mahemacs and Sascs, Zhengzhou Normal Unversy Zhengzhou Cy, Hennan Provnce, 4544, Chna Malng addss: No. 6, Yngca sr., Norh Unversy Cy, Zhengzhou Cy, Henan Provnce, Chna, 4544 cluu6697@sna.com Zhong-guang Fan School of Mahemacs and Sascs, Zhengzhou Normal Unversy. Zhengzhou Cy, Hennan Provnce, 4544, Chna Absrac In hs paper, search sae space srucu propes of normal chan from Ray-Kngh heory on a gven se : If S ( ω) whose probably s 1 a empy se,hen,he normal chan only has jumpng ype rac, a he me, he ransfer funcon mus be he leas and only deermned by desy mar; If S ( ω) whose probably s posve a no empy se, hen,he rac of normal chan s no jumpng ype rac, and s ransfer funcon canno deermne by densy mar. Keywords: Normal chan; Soppng me, Jumpng rac 1 Inroducon [1 3] In he classc boos of some Marov chan,for he ransfer funcon n a gven, he usuall pracce s consrucng canoncal chan on U. he sae space of model chan s very smply. Bu s rac jus has rgh lower sem connuy,

164 L Chen and Zhong-guang Fan and jus eep par of he srong Marov. As applcaon of Ray-ngh heory, he [4 5] normal chan can overcome hese defecs,bu he srucu of sae space may be very comple. So we wll search he propes of he normal chan. 2 Plmnary nowledge [3] Le = {1, 2, L }, P () = ( p()), and P () s he hones ransfer j, j funcon on, Q = ( q ), Q s densy mar of P (), R ( ) s a j, j j solven of P (), s Ray-Kngh gh for P () on, ( U α ) α > s a Ray solven for P (). And P () s Ray sem group for P (), ε s a Bol algebra on +.Le D [, ) = { ωω:[, ) }, F s a Bol algebra on D [, ), Coordnae process of F s { X } ( ), Le Ω denoe he subse { =, } c + = of D [, ). θ s he mappng: R θ : ω() a ω( + ), ω() D [, ),. P s he probably measu of on +. Defnon 1 [6] Le X = ( Ω, F, F, X, θ, P ), X s called normal chan on P (). Defnon 2 For a funcaon on { Ω, F}, : Ω [, ), f for any, he a { } F, hen we called s soppng me. Defnon 3 Le S ( ω) = { s X ( ω) = }, we call S ( ω ) s -consan value se ; he nerval of S ( ω ) s called -nerval of X. s

rac propes of normal chan 165 Defnon 4 If = nf{ s s, X X }, = nf{ s s >, X = X }, s called s s escape me. s called urn me. Defnon 5 For any +, P { = } = or 1, P { = } = or 1. If P { = } = 1, s called absorbng sae; If P { > } = 1, s called say sae; If P { = } = 1, s called gular sae; If P { = } = P { = } = 1, s called nsananeous sae. 3 Relevan heoms heom 1 Le (1) s Regular Sae., hen (2) On P, obey eponenal dsrbuon of parameer q. (3) On P, X and a ndependen. q j (4) If < q <, hen for any j, j, P { X = j} =. q Proof: (1) If s no Regular Sae, hen P { > } = 1, for any > s easy o see { X = } { } and when, { } { = }, so we have ha 1= lm P( ) = lm P { X = } lm P { } = hs s a conradcon sul. So s Regular Sae. (2) Refer o [2]. (3) If q = or, s clearly ha P { = } = 1 or P { = } = 1.,

166 L Chen and Zhong-guang Fan If < q <,for any A, A +, and s>,, P { > + s, X A} = P { >, oθ > s, X o θ A} = { P { oθ > s, X o θ A F}; > } X = { P { > s, X A}; > } = P { > } P { > s, X A} Le s, hen P { >, X A} = P { > } P { X A} So, on P, and X a ndependen. (4) If < q <, for any j, >, from (3) and he srong Marov of X, We ge ha R ( ) e = p ( ) d j = [ ( ) ] j e I X d { j} = [ e I ( X ) d] { j} = [ e [ e I ( X ) d] o θ ] { j} X = [ e [ e I ( X ) d]] { j} = [ e e P [ X = j] d] X = [ e e P( X,{ j}) d] = [ e U ( X,{ j})] = [ e ] [ U ( X,{ j})]

rac propes of normal chan 167 + For any, j, for connuous funcons f () on, so ha f ( ) =, f ( j) = 1, hen = f ( ) = lm U f ( ) lm U (,{ j}). Bu lm U ( j,{ j}) = lm R jj ( ) = 1, So q = R = e U X j 2 lm ( ) lm [ ]lm [ (,{ })] j j q = lm [lm U ( X,{ j})] + q = q P [ X = j] q j ha s P [ X = j] =. q s nsananeous sae(sable sae) f Noe1: From heom 1 we ge ha and only s ha s nsananeous sae(say sae) of normal chan X. heom 2 If ha a b and f q <, hen for any + a <, hen a b, he a a 1, b 1, a 2, b 2 L, so < ; If, 1 b b a + < <, for any, wll have S ( ω) = U [ a ( ω), b ( ω)). For any s <, le ξ (,) s denoe he number of [ a, b ) n [ s,], = 1, 2, L, hen { ξ ( s, )} q( s). Proof: Le

168 L Chen and Zhong-guang Fan a = nf{ u u, X = } 1 b = nf{ u u a, X } 1 1 a = nf{ u u b, X = } + 1 u b = nf{ u u a, X } + 1 u I s easy o now ha a, b, = 1,2, L a { F } soppng me. For any, f a <, because X s rgh connuous, and 1,we have ha u u X a P [ a <, b > a ] = P[ a <, o θ > ] a =, from (2) of heom = [ P [ o θ > F ]; a < ] = [ P [ > ]; a < ] = P [ a < ] a a hen, on { a < }, almos su have a < b, from he srong Marov of X and defne of b, for any, we ge ha = P [ b <, ε >, X [ b, b + ε), X = ] = P { b <, X =, o θ > } b b = [ P [ o θ > F ]; b <, X = ] = [ P [ > ]; b <, X = ] = P { X =, b < } b b b b b So, on { b < }, almos su have X b.for any < s <, s obvous ha ξ (,)} s = ξ (, s) o θ.(fer o [2](12-13)) s

rac propes of normal chan 169 { ξ ( s, )} { ξ (, s) o θ } s = P { X = } { ξ (, s)} q ( s) s For any >, have { ξ (, )} q.so,almos su ha he a only lmed [ a, b ), = 1,2, L on any lmed nerval. hen lm a =, we may ge ha S ( ω) = [ a ( ), b ( )) U ω ω heom 3 If q =, we have: (1) Almos su S ( ω ) does no conan any nerval. (2) Almos su S ( ω ) s own dense se. Proof:(1) I s clear ha S ( ω ) s a oponal se. Le A( ω) = sup{ s s <, s S ( ω)},, ω Ω }, s obvous { A } s monoone ncasng lgh connuous process, and suable { F }. Le B = lm A, hen { B } s s a rgh connuous process of oponal s suable { F }.Se U = {( ω, ) ε >, ( ε, + ε) S ( ω)}, ω Ω, hen U = {( ω, ) B ( ω) < }, so U s he oponal se suable { F }. We use D o denoe he opporuny me U, If P { D < } >, he ess u U { F } soppng me, so ha P { < } >, and on { < },( ω, ( ω)) U, From (2) of heom 1 we have ha P [ < ] = P [ <, X =, ε >, ε ( ε, + ε), X ]

17 L Chen and Zhong-guang Fan P [ <, X =, o θ > ] [ = [ P > ], < ] = hs s a conradcon sul! So P { D < } =.Namely, almos su S ( ω ) U does no conan any nerval. (2) Proof mehods s same o (1). 4 Man Resuls + Proposon1 If \,hen P { = } = 1. Proof: For any, q <, we use ( ) ( ) [, ] a b denoe he -h -nerval of S ( ω ), Le S ( ω) = { u X ( ω) }, S s he me of soppng he f U f vrualsae. For any +, >, + hen, meas S =, f \ f I = f s { S [, ]} { [1 I ( X )] ds} = [1 P { X }] ds = s, we have ha P { = } = 1. Proposon2 If Ω= ( q ) s all sably, For any ω Ω, Le, j j S ( ω) = { s s, ε > ; wll have nfne jumps}, hen S ( ω) s closed se. S ( ω) \ S f ( ω) conan counable pons, and meas S ( ω) almos su s. Proposon3 Le σ( ω) = nf{ ss>, s S ( ω)}, ω Ω, hen σ s { F } Soppng me. Proposon4 he campagn on (, ) \ S ( ω) mnmum ransfer funcon. Proof: For any n and,, j, we denoe: s deermned by he

rac propes of normal chan 171 I s easy ha: h () = P[ X = j.[,] have n jumps] ( n ) j () q h () e δ j j =, ( n) q n 1 s j j h () = e qh ( sdsn ), = 1,2, L P X j σ h p j ( n) mn [ =, < ] = ( ) ( ),,, n j Because (, ) \ S ( ω) s a open se, and all rac on he (, ) \ S ( ω) jumpng ype, for any { F } soppng me, P X j S ω X P a μ mn [ =,[, + ] I ( ) = = ] = ( ) + j So he campagn on (, ) \ S ( ω) funcon. Noe2: If S ( ω) s deermned by he mnmum ransfer whose probably s 1 s empy se, hen, he normal chan only has jumpng ype rac, a he me, he ransfer funcon mus be he leas and be only deermned by densy mar. If S ( ω) whose probably s posve s no empy se, hen he rac of normal chan s no jumpng ype rac, and s ransfer funcon canno deermne by densy mar. Refences [1]ZHONG Kala, Course n probably heory, Shangha scence and echnology Pss.1989 [2]WANG Zun; YANG Xangqun. Brh and deah processes and Marov chan, Bejng Scence Pss.25.55-57 [3]YANG Xangqun. Srucu heory of counable Marov processes.hunan scence and echnology Pss.1981.3-6 [4]HOU Zhengng;GUO qngfeng. Homogeneous counable Marov processes. Bejng scence and echnology Pss.1978. [5]HOU Zhenng. he only cron of Q-processes. Hunan scence and echnology Pss.1982. [6]WU Qunyng;ZHANG Hanjun;HOU Zhenng.An eended brh-deah Q-mar wh nsananeous Sae[J],Chnese J.conemp.Mah,23,24(2)159-168. Receved: November, 212