Path Integral methods for solving stochastic problems. Carson C. Chow, NIH
|
|
- Anissa Poole
- 5 years ago
- Views:
Transcription
1 Path Integral methods for solving stochastic problems Carson C. Chow, NIH
2 Why? Often in neuroscience we run into stochastic ODEs of the form dx dt = f(x)+g(x)η(t) η(t) =0 η(t)η(t ) = δ(t t )
3 Examples Integrate-and-fire neuron with noise dv dt = I g Lv g NL (v)+ Dη(t) Set v v R when v = v T Want to know mean and variance, mean and variance of time to fire, etc.
4 Decision-theory model du dt = f(u)+g(t)+ση(t) u(0) = b Decision is made when u reaches ±θ
5 Population of neurons with noise t a = a + f ( w(x, y)a(y)dy ) + η(x, t) η(x, t) =0 η(x, t)η(x,t ) = δ(x x )δ(t t ) Again interested in mean, variance, correlations, any statistic
6 Solving equations Traditional methods include Ito and Stratonovich calculus or Fokker-Planck equation For most nonlinear stochastic equations, closed form solutions do not exist Need to be solved with perturbation theory, which is nontrivial Path integral and field theory approaches were designed for perturbation theory
7 Generating Function Generating function lets you calculate moments of a probability density P(x) by taking derivatives moments x n = x n P (x)dx generating function x n = 1 Z[0] Z[λ] = e λx = n λ n Z[λ] λ=0 e λx P (x)dx
8 Goals Derive a generating function for moments of a stochastic process Solve for moments perturbatively using diagrammatic methods.
9 Computing Generating functions e.g. Normal or Gaussian distribution P (x) e (x a) 2 2σ 2 Z[λ] = e (x a) 2 2σ 2 +λx dx Most important trick: complete the square
10 (x a)2 2σ 2 x c is given by + λx = A(x x c ) 2 + B ( ) d (x a)2 dx 2σ 2 + λx =0 A = 1 2σ 2 x c = λσ 2 + a B = x2 c 2σ 2 a2 2σ 2 = λ2 σ λa Z[λ] = e (x λσ 2 a) 2 2σ 2 +λa+ λ2 σ 2 2 dx = Z[0]e λa+ λ2 σ 2 2 Z[0] = 2πσ
11 x = d dλ eλa+ λ 2 σ 2 2 x 2 = d2 dλ 2 eλa+ λ 2 σ 2 2 = a λ=0 = a 2 + σ 2 λ=0 x 3 = d3 dλ 3 eλa+ λ 2 σ 2 2 x 4 = d4 dλ 4 eλa+ λ 2 σ 2 2 = a 3 +3aσ 2 λ=0 = a 4 +6a 2 σ 2 +3σ 4 λ=0
12 Cumulant generating function W [λ] = ln Z[λ] C n = dn dλ n W [λ] λ=0 For Gaussian W [λ] =λa (λ2 σ 2 ) + ln Z[0] C 1 = x = a C 2 = x 2 x 2 var(x) =σ 2 C n =0, n > 2
13 Generalize to multi-dimensions Z[λ i ]= dx i e 1 2 Pj,k x jk 1 jk x k+ P j λ jx j K 1 jk (K 1 ) jk i Assume K is symmetric and positive, thus it has full set of orthonormal eigenvectors K ij v α j = ω α v α i v α j v β j = δ αβ j j Expand x k = α c α v α k λ k = α d α v α k x j K jk x k = j c α ω β c β v α j v β k = α,β c α ω β c β δ αβ = α ω α c 2 α j,k α,β
14 Z[λ i ]= = = α i α dc α e P α ( 1 2 ω αc 2 α +d αc α ) dx i e 1 2 Pj,k x jk 1 jk x k+ P j λ jx j dc α e 1 2 ω αc 2 α +d αc α = Z[0] α e 1 2 ω 1 α d2 α Z[0] = (2π det K) n/2 = Z[0]e P jk 1 2 λ jk jk λ k j K 1 ij K jk = δ ik Cumulant generating function W [λ i ] = ln Z[λ i ]
15 Moments i x i = 1 Z[0] i Z[λ i ] λ i λi =0 Example x 2 3x 7 x 36 = 1 Z[0] 2 λ 2 3 λ 7 exp λ 36 ij 1 2 λ ik 1 ij λ j λi =0 = K 3,3 K 7,36 +2K 3,7 K 3,36
16 Wick s Theorem 2s 1 x i = 1 = Z[0] 2s 1 Z[λ i ] λ i all possible pairings λi =0 K i1,i 2 K i2s 1 i 2s x a x b x c x d = x a x b x c x d + x a x b x c x d + x a x b x c x d = K ab K cd + K ad K bc + K ac K bd make all pairings or contractions
17 Continuum limit Z[λ i ]= dx i e 1 2 Pj,k x jk 1 jk x k+ P j λ jx j i Let i t x i x(t) λ i λ(t) K ij K(s, t) dt i Z[λ(t)] = Dx(t)e 1 2 dx i Dx(t) i R x(s)k 1 (s,t)x(t)dsdt+ R λ(t)x(t)dt = Z[0]e R 1 2 λ(s)k(s,t)λ(t)dsdt
18 functional derivative λ i δ δλ(t) δλ(s) δλ(t) = δ(s t) Moments x(t 1 )x(t 2 ) = 1 Z[0] δ δλ(t 1 ) δ δλ(t 2 ) Z[λ(t)] = K(t 1,t 2 ) i x(t i ) = 1 Z[0] i δ δλ(t i ) Z[λ(t)] = all possible contractions K(t i1,t i2 ) K(t i2s 1,t ti2s )
19 Can generalize to higher dimensions x(t) ϕ( t) λ(t) J( t) t R d dt d d t Z[λ( t)] = Dϕe 1 2 R ϕ( s)k 1 ( s, t)ϕ( t)d d sd d t+ R λ( t)ϕ( t)d d t = Z[0]e R 1 2 λ( s)k( s, t)λ( t)d d sd d t
20 Field Theory Consider an arbitrary (infinite dimensional) probability distribution P [ϕ( t)] = e S[ϕ( t)] Z[J(ϕ( t)] = Dϕe S[φ]+J ϕ S is called the action J ϕ = J( t)ϕ( t)d d t e.g. φ 4 theory S[ϕ( t)] = ϕ( t)k 1 ( t, t )ϕ( t )d d td d t + g ϕ 4 ( t)d d t Must compute perturbatively (asymptotic solutions of integrals in infinite dimensions)
21 Application to stochastic differential equations dx Ito interpretation dt = f(x)+g(x)η(t) dx = f(x)dt + g(x)db t Discretization x i+1 x i = f(x i )dt + g(x i )w i dt db t = w t t
22 Probability density P [x i w i ]=δ[x i+1 x i f(x i )dt g(x i )w i dt] = δ(x 1 x 0 f(x 0 )dt g(x 0 )w 0 dt)δ(x2 x 1 Use Fourier transform representation δ(z) = 1 2π e ik iz dk i P [x i w i ]= i dk i 2π e i P i k i(x i+1 x i f(x i )dt g(x i )w i t)
23 P [x i w i ]= i dk i 2π e i P i k i(x i+1 x i f(x i )dt) e ik ig(x i )w i dt wi is Normal or Gaussian N(0,1) with density P [w i ]= 1 2π e 1 2 w2 i P [x i ]= i dw i P [x i w i ]P [w i ] = i dk P i 2π e i i k i(x i+1 x i f(x i )dt) i dw i 2π e ik ig(x i )w i dt e 1 2 w2 i
24 Complete the square P [x i ]= i dk i 2π e Pi (ik i)( x i+1 x i dt f(x i ))dt+ P i 1 2 g2 (x i )(ik i ) 2 dt Make a change of variables ki = ik i Take the continuum limit P [x(t)] = D ke R dt k(ẋ f(x(t))+ 1 2 k 2 g 2 (x(t)) action S[x(t), k(t)] = dt k(ẋ f(x(t)) 1 2 k 2 g 2 (x(t))
25 From SDE to action ẋ = f(x)+g(x)η(t) P [x] = P [x] = P [x] = Dηδ[ẋ f(x) g(x)η(t)]e R η 2 (t)dt DηD ke R k(ẋ f(x))+ kg(x)η(t) η 2 (t)dt D ke R k(ẋ f(x))+ 1 2 k 2 g 2 (x)dt S[x(t), k(t)] = dt k(ẋ f(x(t)) 1 2 k 2 g 2 (x(t))
26 Ornstein-Uhlenbeck ẋ = ax + Dη(t) P [x] = D ke R k(ẋ+ax)+ D 2 k 2 dt G 1 (t t ) = ( d dt + a)δ(t t ) P [x] = D ke R dtdt k(t)g 1 (t t )x(t )+ R dt D 2 k 2 S[x, k] = dtdt k(t)g 1 (t t )x(t ) dt D 2 k 2
27 Generating functional Z[J(t), J(t)] = DxD ke R dtdt kg 1 x+ R D 2 k 2 dt+ R Jxdt+ R J kdt = DxD ke R dtdt kg 1x (1 + µ + 1 2! µ ! µ3 + ) D where µ = 2 k 2 dt + Jxdt + J kdt Expansion in moments of complex Gaussian
28 Z CG [J(t), J(t)] = DxDke R dtdt ikg 1 x+ R ik(t)j(t)dt+ R x(t) J(t)dt Restore k = ik for convenience Discretize and diagonalize dt G 1 x λ(ω)ˆx(ω) Expand in ˆx(ω) Z CG [Ĵ, ˆ J] = k, J, J DˆxDˆke P ω ˆk(ω), Ĵ(ω), ˆ J(ω) iˆkˆxλ+iˆkĵ+ˆx ˆ J = ω dˆxdˆk 2π e iˆkˆxλ+iˆkĵ+ˆx ˆ J
29 = ω = ω dˆxdˆk 2π e iˆk(ˆxλ Ĵ)+ˆx dˆxδ(ˆxλ Ĵ)eˆx ˆ J ˆ J = ω e λ 1 Ĵ ˆ J Transform back to get Z CG [J, J] =e R J(t)G(t,t ) J(t )
30 Propagator ( d dt + a)x = G 1 (t, t )x(t )dt G 1 (t, t )=( d dt + a)δ(t t ) G 1 (t, t )G(t,t )dt = δ(t t ) ( d dt + a)g(t 1,t 2 )=δ(t 1 t 2 ) Green s function
31 Moments ij x(t i ) k(t j ) = ij δ δ J(t i ) δ δj(t j ) er J(t)G(t,t )J(t )dtdt J= J=0 Only surviving moments have equal numbers of x and k Wick s theorem still applies but only for contractions between x and k e.g. x(t 1 )x(t 2 ) k(t 3 ) k(t 4 ) = G(t 1,t 3 )G(t 2,t 4 )+G(t 1,t 4 )G(t 2,t 3 )
32 Back to the OU problem Z[J(t), J(t)] = DxD ke R dtdt kg 1 x+ R D 2 k 2 dt+ R Jxdt+ R J kdt = DxD ke R dtdt kg 1x (1 + µ + 1 2! µ ! µ3 + ) D where µ = 2 k 2 dt + Jxdt + J kdt µ 1 + µ 2 + µ 3
33 Can see that only some combinations of the μ terms will survive i.e. μ 2 μ 3 and μ 1 μ 2 2 Z[J(t), J(t)] = J(t)J(t )x(t) k(t ) dtdt + D 4 J(t) J(t )x(t)x(t ) k 2 (t ) dtdt dt +... x(t 1 )x(t 2 ) = D 2 x(t 1 )x(t 2 ) k 2 (t ) dt = D G(t 1,t )G(t 2,t )dt
34 Feynman diagrams Each term in action can be represented by a diagram Turn perturbation theory into rules for assembling diagrams Diagrams consist of edges and nodes, representing terms in the action
35 Every factor of k is given an outgoing edge Every factor of x is given an incoming edge Convention is time goes from right to left Propagators are lines, other terms are vertices
36 S[x, k] = dtdt k(t)g 1 (t, t )x(t ) dt D 2 k 2 Assign a diagram to each term in the action Propagator G(t, t ) t t Other diagrams are called vertices Vertex D 2 t sign negative from action
37 Feynman rules Surviving terms in perturbation series of generating functional consist of equal numbers of k and x factors For vertices, this means that outgoing edges are joined to incoming edges Propagators are then attached to all edges in all possible ways (accounts for Wick s theorem) Vertices are integrated over
38 Moment rules Moments of x are given by derivatives over J, moments of k are given by derivatives over J Thus moments can also be expressed as diagrams Each factor of each factor of x k is given an outgoing edge and is given an incoming edge Cumulants are connected diagrams
39 Examples Propagator: x(t 1 ) k(t 2 ) = = G(t t,t 2 ) 2nd moment: x(t 1 )x(t 2 ) = =2 G(t 1,t )G(t 2,t ) D 2 dt
40 Doing the integrals ( d dt + a)g(t 1,t 2 )=δ(t 1 t 2 ) G(t 1,t 2 )=e a(t 1 t 2 ) H(t 1 t 2 ) H(t) = { 1 t>0 0 t 0 D G(t 1,t )G(t 2,t )dt = D t2 0 e a(t 1 t ) e a(t 2 t ) dt t 2 >t 1 = D e2a(t 1 t 2 ) e 2a(t 1+t 2 ) 2a = D 2a (1 e 2at ) t 1 = t 2 = t
41 Nonlinear terms ẋ = ax + bx 2 cx 3 + A + Bx + Cx 2 η(t)+x 0 δ(t t 0 ) Using the action formula S[x(t), k(t)] = dt k(ẋ f(x(t)) 1 2 k 2 g 2 (x(t)) S[x(t), k(t)] = dt k(ẋ + ax bx 2 + cx 3 x 0 δ(t t 0 )) k 2 A + Bx + Cx2 2
42 S[x(t), k(t)] = dt k(ẋ + ax bx 2 + cx 3 x 0 δ(t t 0 )) k 2 A + Bx + Cx2 2 Propagator G(t, t ) Vertices x 0 δ(t t 0 ) b c A 2 B 2 C 2
43 x(t 1 )x(t 2 ) = Tree level 2 B 2 x 0 G(t 1,t )G(t 2,t )G(t,t 0 )dt One loop
44 Acknowledgments and References I will be writing a review that will be posted on my blog at sciencehouse.wordpress.com Michael Buice Zinn-Justin, Quantum Field Theory and Critical Phenomena Kardar, Statistical Physics of Fields
arxiv: v2 [nlin.ao] 9 Oct 2012
Path Integral Methods for Stochastic Differential Equations Carson C. Chow and Michael A. Buice Laboratory of Biological Modeling, NIDDK, NIH, Bethesda, MD 20892 (Dated: October 10, 2012) We give a pedagogical
More informationPath Integral Methods for Stochastic Differential Equations
Journal of Mathematical Neuroscience 2015) 5:8 DOI 10.1186/s13408-015-0018-5 RESEARCH OpenAccess Path Integral Methods for Stochastic Differential Equations Carson C. Chow 1 Michael A. Buice 1 Received:
More informationSDE Coefficients. March 4, 2008
SDE Coefficients March 4, 2008 The following is a summary of GARD sections 3.3 and 6., mainly as an overview of the two main approaches to creating a SDE model. Stochastic Differential Equations (SDE)
More informationAdvanced Quantum Mechanics
Advanced Quantum Mechanics Rajdeep Sensarma! sensarma@theory.tifr.res.in Lecture #22 Path Integrals and QM Recap of Last Class Statistical Mechanics and path integrals in imaginary time Imaginary time
More information16. Working with the Langevin and Fokker-Planck equations
16. Working with the Langevin and Fokker-Planck equations In the preceding Lecture, we have shown that given a Langevin equation (LE), it is possible to write down an equivalent Fokker-Planck equation
More informationIntroduction. Stochastic Processes. Will Penny. Stochastic Differential Equations. Stochastic Chain Rule. Expectations.
19th May 2011 Chain Introduction We will Show the relation between stochastic differential equations, Gaussian processes and methods This gives us a formal way of deriving equations for the activity of
More informationThe Path Integral: Basics and Tricks (largely from Zee)
The Path Integral: Basics and Tricks (largely from Zee) Yichen Shi Michaelmas 03 Path-Integral Derivation x f, t f x i, t i x f e H(t f t i) x i. If we chop the path into N intervals of length ɛ, then
More informationPath integrals for classical Markov processes
Path integrals for classical Markov processes Hugo Touchette National Institute for Theoretical Physics (NITheP) Stellenbosch, South Africa Chris Engelbrecht Summer School on Non-Linear Phenomena in Field
More information5 Applying the Fokker-Planck equation
5 Applying the Fokker-Planck equation We begin with one-dimensional examples, keeping g = constant. Recall: the FPE for the Langevin equation with η(t 1 )η(t ) = κδ(t 1 t ) is = f(x) + g(x)η(t) t = x [f(x)p
More informationLecture 4: Ito s Stochastic Calculus and SDE. Seung Yeal Ha Dept of Mathematical Sciences Seoul National University
Lecture 4: Ito s Stochastic Calculus and SDE Seung Yeal Ha Dept of Mathematical Sciences Seoul National University 1 Preliminaries What is Calculus? Integral, Differentiation. Differentiation 2 Integral
More informationStochastic contraction BACS Workshop Chamonix, January 14, 2008
Stochastic contraction BACS Workshop Chamonix, January 14, 2008 Q.-C. Pham N. Tabareau J.-J. Slotine Q.-C. Pham, N. Tabareau, J.-J. Slotine () Stochastic contraction 1 / 19 Why stochastic contraction?
More informationStochastic Differential Equations.
Chapter 3 Stochastic Differential Equations. 3.1 Existence and Uniqueness. One of the ways of constructing a Diffusion process is to solve the stochastic differential equation dx(t) = σ(t, x(t)) dβ(t)
More informationVectors, matrices, eigenvalues and eigenvectors
Vectors, matrices, eigenvalues and eigenvectors 1 ( ) ( ) ( ) Scaling a vector: 0.5V 2 0.5 2 1 = 0.5 = = 1 0.5 1 0.5 ( ) ( ) ( ) ( ) Adding two vectors: V + W 2 1 2 + 1 3 = + = = 1 3 1 + 3 4 ( ) ( ) a
More informationThe concentration of a drug in blood. Exponential decay. Different realizations. Exponential decay with noise. dc(t) dt.
The concentration of a drug in blood Exponential decay C12 concentration 2 4 6 8 1 C12 concentration 2 4 6 8 1 dc(t) dt = µc(t) C(t) = C()e µt 2 4 6 8 1 12 time in minutes 2 4 6 8 1 12 time in minutes
More informationA path integral approach to the Langevin equation
A path integral approach to the Langevin equation - Ashok Das Reference: A path integral approach to the Langevin equation, A. Das, S. Panda and J. R. L. Santos, arxiv:1411.0256 (to be published in Int.
More informationPath Intergal. 1 Introduction. 2 Derivation From Schrödinger Equation. Shoichi Midorikawa
Path Intergal Shoichi Midorikawa 1 Introduction The Feynman path integral1 is one of the formalism to solve the Schrödinger equation. However this approach is not peculiar to quantum mechanics, and M.
More informationGaussian processes for inference in stochastic differential equations
Gaussian processes for inference in stochastic differential equations Manfred Opper, AI group, TU Berlin November 6, 2017 Manfred Opper, AI group, TU Berlin (TU Berlin) inference in SDE November 6, 2017
More informationSession 1: Probability and Markov chains
Session 1: Probability and Markov chains 1. Probability distributions and densities. 2. Relevant distributions. 3. Change of variable. 4. Stochastic processes. 5. The Markov property. 6. Markov finite
More informationMathematical Methods for Neurosciences. ENS - Master MVA Paris 6 - Master Maths-Bio ( )
Mathematical Methods for Neurosciences. ENS - Master MVA Paris 6 - Master Maths-Bio (2014-2015) Etienne Tanré - Olivier Faugeras INRIA - Team Tosca November 26th, 2014 E. Tanré (INRIA - Team Tosca) Mathematical
More informationHarnack Inequalities and Applications for Stochastic Equations
p. 1/32 Harnack Inequalities and Applications for Stochastic Equations PhD Thesis Defense Shun-Xiang Ouyang Under the Supervision of Prof. Michael Röckner & Prof. Feng-Yu Wang March 6, 29 p. 2/32 Outline
More informationContinuous Random Variables
1 / 24 Continuous Random Variables Saravanan Vijayakumaran sarva@ee.iitb.ac.in Department of Electrical Engineering Indian Institute of Technology Bombay February 27, 2013 2 / 24 Continuous Random Variables
More informationMath 4381 / 6378 Symmetry Analysis
Math 438 / 6378 Smmetr Analsis Elementar ODE Review First Order Equations Ordinar differential equations of the form = F(x, ( are called first order ordinar differential equations. There are a variet of
More informationMATH 56A SPRING 2008 STOCHASTIC PROCESSES 197
MATH 56A SPRING 8 STOCHASTIC PROCESSES 197 9.3. Itô s formula. First I stated the theorem. Then I did a simple example to make sure we understand what it says. Then I proved it. The key point is Lévy s
More informationLANGEVIN EQUATION AND THERMODYNAMICS
LANGEVIN EQUATION AND THERMODYNAMICS RELATING STOCHASTIC DYNAMICS WITH THERMODYNAMIC LAWS November 10, 2017 1 / 20 MOTIVATION There are at least three levels of description of classical dynamics: thermodynamic,
More informationHomogenization with stochastic differential equations
Homogenization with stochastic differential equations Scott Hottovy shottovy@math.arizona.edu University of Arizona Program in Applied Mathematics October 12, 2011 Modeling with SDE Use SDE to model system
More informationp 1 ( Y p dp) 1/p ( X p dp) 1 1 p
Doob s inequality Let X(t) be a right continuous submartingale with respect to F(t), t 1 P(sup s t X(s) λ) 1 λ {sup s t X(s) λ} X + (t)dp 2 For 1 < p
More informationBernardo D Auria Stochastic Processes /10. Notes. Abril 13 th, 2010
1 Stochastic Calculus Notes Abril 13 th, 1 As we have seen in previous lessons, the stochastic integral with respect to the Brownian motion shows a behavior different from the classical Riemann-Stieltjes
More informationState Space Representation of Gaussian Processes
State Space Representation of Gaussian Processes Simo Särkkä Department of Biomedical Engineering and Computational Science (BECS) Aalto University, Espoo, Finland June 12th, 2013 Simo Särkkä (Aalto University)
More informationGaussian integrals and Feynman diagrams. February 28
Gaussian integrals and Feynman diagrams February 28 Introduction A mathematician is one to whom the equality e x2 2 dx = 2π is as obvious as that twice two makes four is to you. Lord W.T. Kelvin to his
More information1 Probability : Worked Examples
1 Probability : Worked Examples 1) The information entropy of a distribution {p n } is defined as S = n p n log 2 p n, where n ranges over all possible configurations of a given physical system and p n
More informationStability of Stochastic Differential Equations
Lyapunov stability theory for ODEs s Stability of Stochastic Differential Equations Part 1: Introduction Department of Mathematics and Statistics University of Strathclyde Glasgow, G1 1XH December 2010
More informationHandbook of Stochastic Methods
C. W. Gardiner Handbook of Stochastic Methods for Physics, Chemistry and the Natural Sciences Third Edition With 30 Figures Springer Contents 1. A Historical Introduction 1 1.1 Motivation I 1.2 Some Historical
More informationChapter 3, 4 Random Variables ENCS Probability and Stochastic Processes. Concordia University
Chapter 3, 4 Random Variables ENCS6161 - Probability and Stochastic Processes Concordia University ENCS6161 p.1/47 The Notion of a Random Variable A random variable X is a function that assigns a real
More informationx 3y 2z = 6 1.2) 2x 4y 3z = 8 3x + 6y + 8z = 5 x + 3y 2z + 5t = 4 1.5) 2x + 8y z + 9t = 9 3x + 5y 12z + 17t = 7
Linear Algebra and its Applications-Lab 1 1) Use Gaussian elimination to solve the following systems x 1 + x 2 2x 3 + 4x 4 = 5 1.1) 2x 1 + 2x 2 3x 3 + x 4 = 3 3x 1 + 3x 2 4x 3 2x 4 = 1 x + y + 2z = 4 1.4)
More informationADVANCED QUANTUM FIELD THEORY
Imperial College London MSc EXAMINATION May 2016 This paper is also taken for the relevant Examination for the Associateship ADVANCED QUANTUM FIELD THEORY For Students in Quantum Fields and Fundamental
More informationA Class of Fractional Stochastic Differential Equations
Vietnam Journal of Mathematics 36:38) 71 79 Vietnam Journal of MATHEMATICS VAST 8 A Class of Fractional Stochastic Differential Equations Nguyen Tien Dung Department of Mathematics, Vietnam National University,
More informationKalman Filter. Predict: Update: x k k 1 = F k x k 1 k 1 + B k u k P k k 1 = F k P k 1 k 1 F T k + Q
Kalman Filter Kalman Filter Predict: x k k 1 = F k x k 1 k 1 + B k u k P k k 1 = F k P k 1 k 1 F T k + Q Update: K = P k k 1 Hk T (H k P k k 1 Hk T + R) 1 x k k = x k k 1 + K(z k H k x k k 1 ) P k k =(I
More informationLecture 6: Bayesian Inference in SDE Models
Lecture 6: Bayesian Inference in SDE Models Bayesian Filtering and Smoothing Point of View Simo Särkkä Aalto University Simo Särkkä (Aalto) Lecture 6: Bayesian Inference in SDEs 1 / 45 Contents 1 SDEs
More informationTEST CODE: MMA (Objective type) 2015 SYLLABUS
TEST CODE: MMA (Objective type) 2015 SYLLABUS Analytical Reasoning Algebra Arithmetic, geometric and harmonic progression. Continued fractions. Elementary combinatorics: Permutations and combinations,
More informationNonlinear Filtering Revisited: a Spectral Approach, II
Nonlinear Filtering Revisited: a Spectral Approach, II Sergey V. Lototsky Center for Applied Mathematical Sciences University of Southern California Los Angeles, CA 90089-3 sergey@cams-00.usc.edu Remijigus
More informationStochastic Calculus February 11, / 33
Martingale Transform M n martingale with respect to F n, n =, 1, 2,... σ n F n (σ M) n = n 1 i= σ i(m i+1 M i ) is a Martingale E[(σ M) n F n 1 ] n 1 = E[ σ i (M i+1 M i ) F n 1 ] i= n 2 = σ i (M i+1 M
More informationFormulas for probability theory and linear models SF2941
Formulas for probability theory and linear models SF2941 These pages + Appendix 2 of Gut) are permitted as assistance at the exam. 11 maj 2008 Selected formulae of probability Bivariate probability Transforms
More informationChapter Two: Numerical Methods for Elliptic PDEs. 1 Finite Difference Methods for Elliptic PDEs
Chapter Two: Numerical Methods for Elliptic PDEs Finite Difference Methods for Elliptic PDEs.. Finite difference scheme. We consider a simple example u := subject to Dirichlet boundary conditions ( ) u
More informationQualification Exam: Mathematical Methods
Qualification Exam: Mathematical Methods Name:, QEID#41534189: August, 218 Qualification Exam QEID#41534189 2 1 Mathematical Methods I Problem 1. ID:MM-1-2 Solve the differential equation dy + y = sin
More informationStochastic Modelling in Climate Science
Stochastic Modelling in Climate Science David Kelly Mathematics Department UNC Chapel Hill dtbkelly@gmail.com November 16, 2013 David Kelly (UNC) Stochastic Climate November 16, 2013 1 / 36 Why use stochastic
More informationTEST CODE: MIII (Objective type) 2010 SYLLABUS
TEST CODE: MIII (Objective type) 200 SYLLABUS Algebra Permutations and combinations. Binomial theorem. Theory of equations. Inequalities. Complex numbers and De Moivre s theorem. Elementary set theory.
More informationChapter 6 - Random Processes
EE385 Class Notes //04 John Stensby Chapter 6 - Random Processes Recall that a random variable X is a mapping between the sample space S and the extended real line R +. That is, X : S R +. A random process
More informationMath Camp II. Calculus. Yiqing Xu. August 27, 2014 MIT
Math Camp II Calculus Yiqing Xu MIT August 27, 2014 1 Sequence and Limit 2 Derivatives 3 OLS Asymptotics 4 Integrals Sequence Definition A sequence {y n } = {y 1, y 2, y 3,..., y n } is an ordered set
More informationMultiple Random Variables
Multiple Random Variables Joint Probability Density Let X and Y be two random variables. Their joint distribution function is F ( XY x, y) P X x Y y. F XY ( ) 1, < x
More informationOptimisation and optimal control 1: Tutorial solutions (calculus of variations) t 3
Optimisation and optimal control : Tutorial solutions calculus of variations) Question i) Here we get x d ) = d ) = 0 Hence ṫ = ct 3 where c is an arbitrary constant. Integrating, xt) = at 4 + b where
More informationStochastic Differential Equations
Chapter 5 Stochastic Differential Equations We would like to introduce stochastic ODE s without going first through the machinery of stochastic integrals. 5.1 Itô Integrals and Itô Differential Equations
More informationRandom Averaging. Eli Ben-Naim Los Alamos National Laboratory. Paul Krapivsky (Boston University) John Machta (University of Massachusetts)
Random Averaging Eli Ben-Naim Los Alamos National Laboratory Paul Krapivsky (Boston University) John Machta (University of Massachusetts) Talk, papers available from: http://cnls.lanl.gov/~ebn Plan I.
More informationProbability and Distributions
Probability and Distributions What is a statistical model? A statistical model is a set of assumptions by which the hypothetical population distribution of data is inferred. It is typically postulated
More informationDerivation of Itô SDE and Relationship to ODE and CTMC Models
Derivation of Itô SDE and Relationship to ODE and CTMC Models Biomathematics II April 23, 2015 Linda J. S. Allen Texas Tech University TTU 1 Euler-Maruyama Method for Numerical Solution of an Itô SDE dx(t)
More informationOn a class of stochastic differential equations in a financial network model
1 On a class of stochastic differential equations in a financial network model Tomoyuki Ichiba Department of Statistics & Applied Probability, Center for Financial Mathematics and Actuarial Research, University
More informationRobust control and applications in economic theory
Robust control and applications in economic theory In honour of Professor Emeritus Grigoris Kalogeropoulos on the occasion of his retirement A. N. Yannacopoulos Department of Statistics AUEB 24 May 2013
More informationLangevin Methods. Burkhard Dünweg Max Planck Institute for Polymer Research Ackermannweg 10 D Mainz Germany
Langevin Methods Burkhard Dünweg Max Planck Institute for Polymer Research Ackermannweg 1 D 55128 Mainz Germany Motivation Original idea: Fast and slow degrees of freedom Example: Brownian motion Replace
More information04. Random Variables: Concepts
University of Rhode Island DigitalCommons@URI Nonequilibrium Statistical Physics Physics Course Materials 215 4. Random Variables: Concepts Gerhard Müller University of Rhode Island, gmuller@uri.edu Creative
More information2012 NCTS Workshop on Dynamical Systems
Barbara Gentz gentz@math.uni-bielefeld.de http://www.math.uni-bielefeld.de/ gentz 2012 NCTS Workshop on Dynamical Systems National Center for Theoretical Sciences, National Tsing-Hua University Hsinchu,
More informationIntroduction to numerical simulations for Stochastic ODEs
Introduction to numerical simulations for Stochastic ODEs Xingye Kan Illinois Institute of Technology Department of Applied Mathematics Chicago, IL 60616 August 9, 2010 Outline 1 Preliminaries 2 Numerical
More informationEEL 5544 Noise in Linear Systems Lecture 30. X (s) = E [ e sx] f X (x)e sx dx. Moments can be found from the Laplace transform as
L30-1 EEL 5544 Noise in Linear Systems Lecture 30 OTHER TRANSFORMS For a continuous, nonnegative RV X, the Laplace transform of X is X (s) = E [ e sx] = 0 f X (x)e sx dx. For a nonnegative RV, the Laplace
More informationIntroduction to nonequilibrium physics
Introduction to nonequilibrium physics Jae Dong Noh December 18, 2016 Preface This is a note for the lecture given in the 2016 KIAS-SNU Physics Winter Camp which is held at KIAS in December 17 23, 2016.
More informationLinear Response and Onsager Reciprocal Relations
Linear Response and Onsager Reciprocal Relations Amir Bar January 1, 013 Based on Kittel, Elementary statistical physics, chapters 33-34; Kubo,Toda and Hashitsume, Statistical Physics II, chapter 1; and
More informationLecture 6: Multiple Model Filtering, Particle Filtering and Other Approximations
Lecture 6: Multiple Model Filtering, Particle Filtering and Other Approximations Department of Biomedical Engineering and Computational Science Aalto University April 28, 2010 Contents 1 Multiple Model
More informationStochastic continuity equation and related processes
Stochastic continuity equation and related processes Gabriele Bassi c Armando Bazzani a Helmut Mais b Giorgio Turchetti a a Dept. of Physics Univ. of Bologna, INFN Sezione di Bologna, ITALY b DESY, Hamburg,
More informationReview: control, feedback, etc. Today s topic: state-space models of systems; linearization
Plan of the Lecture Review: control, feedback, etc Today s topic: state-space models of systems; linearization Goal: a general framework that encompasses all examples of interest Once we have mastered
More informationStochastic Viral Dynamics with Beddington-DeAngelis Functional Response
Stochastic Viral Dynamics with Beddington-DeAngelis Functional Response Junyi Tu, Yuncheng You University of South Florida, USA you@mail.usf.edu IMA Workshop in Memory of George R. Sell June 016 Outline
More informationREVIEW REVIEW. Quantum Field Theory II
Quantum Field Theory II PHYS-P 622 Radovan Dermisek, Indiana University Notes based on: M. Srednicki, Quantum Field Theory Chapters: 13, 14, 16-21, 26-28, 51, 52, 61-68, 44, 53, 69-74, 30-32, 84-86, 75,
More informationQuantum Field Theory II
Quantum Field Theory II PHYS-P 622 Radovan Dermisek, Indiana University Notes based on: M. Srednicki, Quantum Field Theory Chapters: 13, 14, 16-21, 26-28, 51, 52, 61-68, 44, 53, 69-74, 30-32, 84-86, 75,
More informationFisher Information in Gaussian Graphical Models
Fisher Information in Gaussian Graphical Models Jason K. Johnson September 21, 2006 Abstract This note summarizes various derivations, formulas and computational algorithms relevant to the Fisher information
More informationReview of scalar field theory. Srednicki 5, 9, 10
Review of scalar field theory Srednicki 5, 9, 10 2 The LSZ reduction formula based on S-5 In order to describe scattering experiments we need to construct appropriate initial and final states and calculate
More informationChapter 3: Random Variables 1
Chapter 3: Random Variables 1 Yunghsiang S. Han Graduate Institute of Communication Engineering, National Taipei University Taiwan E-mail: yshan@mail.ntpu.edu.tw 1 Modified from the lecture notes by Prof.
More informationImplicit sampling for particle filters. Alexandre Chorin, Mathias Morzfeld, Xuemin Tu, Ethan Atkins
0/20 Implicit sampling for particle filters Alexandre Chorin, Mathias Morzfeld, Xuemin Tu, Ethan Atkins University of California at Berkeley 2/20 Example: Try to find people in a boat in the middle of
More informationRandom Eigenvalue Problems Revisited
Random Eigenvalue Problems Revisited S Adhikari Department of Aerospace Engineering, University of Bristol, Bristol, U.K. Email: S.Adhikari@bristol.ac.uk URL: http://www.aer.bris.ac.uk/contact/academic/adhikari/home.html
More informationCDS 101/110a: Lecture 2.1 Dynamic Behavior
CDS 11/11a: Lecture 2.1 Dynamic Behavior Richard M. Murray 6 October 28 Goals: Learn to use phase portraits to visualize behavior of dynamical systems Understand different types of stability for an equilibrium
More information2.1 Calculation of the ground state energy via path integral
Chapter 2 Instantons in Quantum Mechanics Before describing the instantons and their effects in non-abelian gauge theories, it is instructive to become familiar with the relevant ideas in the context of
More informationSrednicki Chapter 9. QFT Problems & Solutions. A. George. August 21, Srednicki 9.1. State and justify the symmetry factors in figure 9.
Srednicki Chapter 9 QFT Problems & Solutions A. George August 2, 22 Srednicki 9.. State and justify the symmetry factors in figure 9.3 Swapping the sources is the same thing as swapping the ends of the
More informationSDS 321: Introduction to Probability and Statistics
SDS 321: Introduction to Probability and Statistics Lecture 14: Continuous random variables Purnamrita Sarkar Department of Statistics and Data Science The University of Texas at Austin www.cs.cmu.edu/
More informationDiscretization of SDEs: Euler Methods and Beyond
Discretization of SDEs: Euler Methods and Beyond 09-26-2006 / PRisMa 2006 Workshop Outline Introduction 1 Introduction Motivation Stochastic Differential Equations 2 The Time Discretization of SDEs Monte-Carlo
More informationFunctional differentiation
Functional differentiation March 22, 2016 1 Functions vs. functionals What distinguishes a functional such as the action S [x (t] from a function f (x (t, is that f (x (t is a number for each value of
More informationChapter 2. Random Variable. Define single random variables in terms of their PDF and CDF, and calculate moments such as the mean and variance.
Chapter 2 Random Variable CLO2 Define single random variables in terms of their PDF and CDF, and calculate moments such as the mean and variance. 1 1. Introduction In Chapter 1, we introduced the concept
More informationStochastic (intermittent) Spikes and Strong Noise Limit of SDEs.
Stochastic (intermittent) Spikes and Strong Noise Limit of SDEs. D. Bernard in collaboration with M. Bauer and (in part) A. Tilloy. IPAM-UCLA, Jan 2017. 1 Strong Noise Limit of (some) SDEs Stochastic differential
More informationMath 108b: Notes on the Spectral Theorem
Math 108b: Notes on the Spectral Theorem From section 6.3, we know that every linear operator T on a finite dimensional inner product space V has an adjoint. (T is defined as the unique linear operator
More information1 Elementary probability
1 Elementary probability Problem 1.1 (*) A coin is thrown several times. Find the probability, that at the n-th experiment: (a) Head appears for the first time (b) Head and Tail have appeared equal number
More informationSection 4: The Quantum Scalar Field
Physics 8.323 Section 4: The Quantum Scalar Field February 2012 c 2012 W. Taylor 8.323 Section 4: Quantum scalar field 1 / 19 4.1 Canonical Quantization Free scalar field equation (Klein-Gordon) ( µ µ
More informationThe Wiener Itô Chaos Expansion
1 The Wiener Itô Chaos Expansion The celebrated Wiener Itô chaos expansion is fundamental in stochastic analysis. In particular, it plays a crucial role in the Malliavin calculus as it is presented in
More informationPROBABILITY: LIMIT THEOREMS II, SPRING HOMEWORK PROBLEMS
PROBABILITY: LIMIT THEOREMS II, SPRING 218. HOMEWORK PROBLEMS PROF. YURI BAKHTIN Instructions. You are allowed to work on solutions in groups, but you are required to write up solutions on your own. Please
More informationStochastic Calculus Made Easy
Stochastic Calculus Made Easy Most of us know how standard Calculus works. We know how to differentiate, how to integrate etc. But stochastic calculus is a totally different beast to tackle; we are trying
More informationkg meter ii) Note the dimensions of ρ τ are kg 2 velocity 2 meter = 1 sec 2 We will interpret this velocity in upcoming slides.
II. Generalizing the 1-dimensional wave equation First generalize the notation. i) "q" has meant transverse deflection of the string. Replace q Ψ, where Ψ may indicate other properties of the medium that
More informationGaussian Process Approximations of Stochastic Differential Equations
Gaussian Process Approximations of Stochastic Differential Equations Cédric Archambeau Dan Cawford Manfred Opper John Shawe-Taylor May, 2006 1 Introduction Some of the most complex models routinely run
More informationA numerical method for solving uncertain differential equations
Journal of Intelligent & Fuzzy Systems 25 (213 825 832 DOI:1.3233/IFS-12688 IOS Press 825 A numerical method for solving uncertain differential equations Kai Yao a and Xiaowei Chen b, a Department of Mathematical
More informationLAN property for ergodic jump-diffusion processes with discrete observations
LAN property for ergodic jump-diffusion processes with discrete observations Eulalia Nualart (Universitat Pompeu Fabra, Barcelona) joint work with Arturo Kohatsu-Higa (Ritsumeikan University, Japan) &
More informationB8.3 Mathematical Models for Financial Derivatives. Hilary Term Solution Sheet 2
B8.3 Mathematical Models for Financial Derivatives Hilary Term 18 Solution Sheet In the following W t ) t denotes a standard Brownian motion and t > denotes time. A partition π of the interval, t is a
More informationMath 250B Final Exam Review Session Spring 2015 SOLUTIONS
Math 5B Final Exam Review Session Spring 5 SOLUTIONS Problem Solve x x + y + 54te 3t and y x + 4y + 9e 3t λ SOLUTION: We have det(a λi) if and only if if and 4 λ only if λ 3λ This means that the eigenvalues
More informationUNIT 3 INTEGRATION 3.0 INTRODUCTION 3.1 OBJECTIVES. Structure
Calculus UNIT 3 INTEGRATION Structure 3.0 Introduction 3.1 Objectives 3.2 Basic Integration Rules 3.3 Integration by Substitution 3.4 Integration of Rational Functions 3.5 Integration by Parts 3.6 Answers
More informationFourier transforms, Generalised functions and Greens functions
Fourier transforms, Generalised functions and Greens functions T. Johnson 2015-01-23 Electromagnetic Processes In Dispersive Media, Lecture 2 - T. Johnson 1 Motivation A big part of this course concerns
More informationElementary ODE Review
Elementary ODE Review First Order ODEs First Order Equations Ordinary differential equations of the fm y F(x, y) () are called first der dinary differential equations. There are a variety of techniques
More informationRigorous Functional Integration with Applications to Nelson s and the Pauli-Fierz Model
Rigorous Functional Integration with Applications to Nelson s and the Pauli-Fierz Model József Lőrinczi Zentrum Mathematik, Technische Universität München and School of Mathematics, Loughborough University
More informationMortality Surface by Means of Continuous Time Cohort Models
Mortality Surface by Means of Continuous Time Cohort Models Petar Jevtić, Elisa Luciano and Elena Vigna Longevity Eight 2012, Waterloo, Canada, 7-8 September 2012 Outline 1 Introduction Model construction
More informationUNIVERSITY OF SOUTHAMPTON. A foreign language dictionary (paper version) is permitted provided it contains no notes, additions or annotations.
UNIVERSITY OF SOUTHAMPTON MATH055W SEMESTER EXAMINATION 03/4 MATHEMATICS FOR ELECTRONIC & ELECTRICAL ENGINEERING Duration: 0 min Solutions Only University approved calculators may be used. A foreign language
More information