Saturday, April 25, 2015

mathematical-neuroscience.com The generating functional be an infinite dimensional generalization for the familiar generating function for a single random variable.

single random variable.

[PDF]Path Integral Methods for Stochastic Differential Equations
www.mathematical-neuroscience.com/content/.../s13408-015-0018-5.pd...
by CC Chow - ‎2015
Feb 12, 2015 - closed form and must be tackled approximately using methods that include ... Although Wiener introduced path integrals to study stochastic ... agrams as a tool for carrying out perturbative expansions and introduces the “loop.

2 Moment Generating Functionals
The strategy of path integral methods is to derive a generating function or functional
for the moments and response functions for SDEs. The generating functional will
be an infinite dimensional generalization for the familiar generating function for a
single random variable. In this section we review moment generating functions and
show how they can be generalized to functional distributions.
Consider a probability density function (PDF) P(x) for a single real variable x.
The moments of the PDF are given by
xn = xnP(x)dx
and can be obtained directly by taking derivatives of the generating function
Z(J ) = eJx = eJxP(x)dx,
where J is a complex parameter, with
xn = 1
Z(0)
dn
dJn
Z(J )

J=0
.
Note that in explicitly including Z(0) we are allowing for the possibility that P(x)
is not normalized. This freedom will be convenient especially when we apply perturbation
theory. The generating function is



The analogy between stochastic systems and quantum theory, where path integrals
are commonly used, is seen by transforming the time coordinates in the path integrals
via t →

−1t . When the field ϕ is a function of a single variable t , then this would
be analogous to single particle quantum mechanics where the quantum amplitude can
be expressed in terms of a path integral over a configuration variable φ(t). When the
field is a function of two or more variables ϕ(
r, t), then this is analogous to quantum
field theory, where the



3 Application to SDE
Building on the previous section, here we derive a generating functional for SDEs.
Consider a Langevin equation,
dx
dt
= f (x, t)+g(x, t)η(t), (3)
on the domain t ∈ [0,T ] with initial condition x(t0) = y. The stochastic forcing term
η(t) obeys η(t) = 0 and η(t)η(t

) = δ(t −t

). Equation (3) is to be interpreted as
the Ito stochastic differential equation
dx = f (xt , t)dt + g(xt , t)dWt , (4)
where Wt is a Wiener process (i.e. Gaussian white noise), and xt is the value of x
at time t . We show how to generalize to other stochastic processes later. Functions
f and g are assumed to obey all the properties required for an Ito SDE to be well
posed [31]. In particular, the stochastic increment dWt does not depend on f (xt , t)
or g(xt , t) (i.e. xt is adapted to the filtration generated by the noise). The choice
between Ito and Stratonovich conventions amounts to a choice of the measure for
the path integrals, which will be manifested in a condition on the linear response or
“propagator” that we introduce below.


生成方程组
cmp

来自: cmp(const void*, const void*) 2014-03-22 11:24:32

  • Everett

    Everett (╮(╯▽╰)╭ ~(= ̄ U  ̄=)~) 2014-03-22 13:40:51

    线性方程的话直接构造矩阵求解就可以了。矩阵的构造方法是 SparseArray
  • cmp

    cmp (const void*, const void*) 2014-03-22 14:03:37

    线性方程的话直接构造矩阵求解就可以了。矩阵的构造方法是 SparseArray 线性方程的话直接构造矩阵求解就可以了。矩阵的构造方法是 SparseArray Everett
    嗯不是线性的… 是 J sum{ -cos θ_i sin θ_j - Δ sin θ_i cos θ_j} - h sin θ_i == 0 这样的。

    我想了想用了 Table + 用于处理下标的自定义的函数,貌似可行。然后用 NSolve 一下。

    还剩几个问题。像这样一大群待求变量,用 Mathematica 怎么处理比较好呢?用下标,List 还是什么?

    另外,像这种自定义函数,如何封装比较好? Module 么?
  • Everett

    Everett (╮(╯▽╰)╭ ~(= ̄ U  ̄=)~) 2014-03-22 14:37:23

    变量本身是下标的函数,比如θ[i]
  • Everett

    Everett (╮(╯▽╰)╭ ~(= ̄ U  ̄=)~) 2014-03-22 14:38:29

    嗯不是线性的… 是 J sum{ -cos θ_i sin θ_j - Δ sin θ_i cos θ_j} - h sin θ_i == 0 这样 嗯不是线性的… 是 J sum{ -cos θ_i sin θ_j - Δ sin θ_i cos θ_j} - h sin θ_i == 0 这样的。 我想了想用了 Table + 用于处理下标的自定义的函数,貌似可行。然后用 NSolve 一下。 还剩几个问题。像这样一大群待求变量,用 Mathematica 怎么处理比较好呢?用下标,List 还是什么? 另外,像这种自定义函数,如何封装比较好? Module 么? ... cmp
    未必需要定义处理下标的函数,一般是用模式匹配 {i_,j_}/;Mod[j-i,L]==1
  • cmp

    cmp (const void*, const void*) 2014-03-28 12:53:50

    未必需要定义处理下标的函数,一般是用模式匹配 {i_,j_}/;Mod[j-i,L]==1 未必需要定义处理下标的函数,一般是用模式匹配 {i_,j_}/;Mod[j-i,L]==1 Everett
    我终于解决了这些问题!

    我最早想要的是最小化一个 hamiltonian. 后来发现还是用 NMinimize 比较好, 用 NSolve 经常跑不出来.

    然后格点不要太大, 老板说 20~30个变量就好了. 太大了最优化算法不收敛.

No comments:

Post a Comment