Questions tagged [stochastic-filtering]

Stochastic filtering deals with the problem of finding the best estimate for a signal, given a noisy or incomplete observation.

Filter by
Sorted by
Tagged with
0 votes
0 answers
10 views

Is the Kalman Filter computationally optimal for Kalman filtering?

Kalman filtering is known to be a recursive process that minimizes mean square error in linear problems. My question is: has anybody shown that this algorithm is computationally optimal, i.e. that you ...
user avatar
1 vote
0 answers
26 views

What's the autocovariance of 2 concurrent hidden states of a stochastic gaussian walk, using the Kalman filter?

A latent variable evolves as $x_t = x_{t-1} + e_t$ where $e_t$ has a gaussian distribution with 0 mean and a variance which defines the volatility of the overall process. Let $o_t$ be the noisy ...
user avatar
0 votes
0 answers
86 views

Stochastic filtering with time delayed observation

Let $X_t$ be a suitably nice real valued Markov process. The primary two cases I have in mind are a finite state space Markov process, and a Ito diffusion. Define the observation process $Y_t$ by $$...
user avatar
  • 1,519
0 votes
1 answer
114 views

What are the optimal times to sample a process?

Let $X$ be a one dimensional Ito diffusion given by $$X_t = b \,W_t$$ where $b$ is a constant, and $W$ is a standard Brownian motion. Let $B$ be another Brownian motion independent of $W$, and define ...
user avatar
  • 1,519
1 vote
1 answer
161 views

Filtering the period and amplitude of a sine wave corrupted by noise

Let $W$ be a standard Brownian motion and $\mathcal F_t$ its natural filtration. Suppose $\theta, A$ are positive $L^1$ random variables independent of $\mathcal F_t$. Let $Y_t$ be the process $$Y_t :=...
user avatar
  • 1,519
1 vote
0 answers
60 views

When enlarging a filtration makes a stochastic processes into a solution to an SDE

Let $n$ be a positive integer and let $(Y_t)_{t\in [0,1]}$ on $\mathbb{R}^n$ be a stochastic process defined on a filtered probability space $(\Omega,\mathcal{F},(\mathcal{F}_t)_{t\in [0,1]},\mathbb{P}...
user avatar
  • 4,997
0 votes
0 answers
20 views

Explicit bound computation for Kalman filter covariance $\|P_{k|k}\|$ as a function of $\|P_0\|$

My Goal: Given and initial covariance $P_0$, a Kalman filter updates covariance $P_{k|k}$ according to a nonlinear update equation. I am looking for a bound $B=B(\|P_0\|)$ such that $\|P_{k|k}\|\leq ...
user avatar
0 votes
0 answers
27 views

Estimator for last value of an Ornstein-Uhlenbeck process given noisy measurement

I have a (discrete) Ornstein Uhlenbeck process $B_{n+1} = aB_n + X_n$, of which I have a noisy measurement $M_n = B_n + Y_n$ (where $Y$ is a white noise), of which I know the parameters. I am looking ...
user avatar
1 vote
0 answers
45 views

Conditions ensuring that conditional law of a process belongs to a given exponential family

Let $(X_t,Y_t)_{t\geq 0}$ be a pair of $\mathbb{R}^n$-(resp. $\mathbb{R}^m$)-valued stochastic processes on a filtered probability space $(\Omega,\mathcal{F},(\mathcal{F}_t)_{t\geq 0},\mathbb{P})$, ...
user avatar
2 votes
1 answer
64 views

Preservation of the Markov Property under Conditioning

Let $(X_t,Z_t)_t$ be an $\mathbb{R}^{n}\times \mathbb{R}^m$-valued time-homogeneous Markov process on a filtered probability space $(\Omega,\mathcal{F},(\mathcal{F}_t)_t,\mathbb{P})$ with transition ...
user avatar
1 vote
0 answers
68 views

If $(\alpha_t)$ is $\mathbb{F}^X$-progressive for a continuous process $(X_t)$, can we write $\alpha_t = \tilde{\alpha}(t,X)$?

Let $X = (X_t)_{t \geq 0}$ be a continuous, real-valued process defined on some probability space $(\Omega,\mathcal{F},P)$, and let $\mathbb{F}^X = (\mathcal{F}_{t}^X)_{t \geq 0}$ be the filtration ...
user avatar
  • 267
2 votes
0 answers
59 views

The optimality of Kalman filtering

It is known that the Kalman filter estimates the state of the following system recursively. $$x_{k+1}=Ax_k+w_k, \ \ w_k \sim \mathcal{N}(0,Q)$$ $$y_k=Cx_k+v_k, \ \ v_k \sim \mathcal{N}(0,W)$$ In the ...
user avatar
1 vote
1 answer
121 views

Onsager--Machlup functional as the density across a mesh of discrete points

It is known that the ratio of the probability of infinitesimal tubes around paths of Itō diffusion processes converges to the Onsager--Machlup (OM) functional. I wonder whether the ratio of the joint ...
user avatar
0 votes
0 answers
78 views

Parseval's equivalent of Norm that includes a Projection matrix

I need to optimize the norm, ${\bf x}^H {\bf P}_{\bf B} {\bf x} $, where, ${\bf P}_{\bf B} = {\bf B}^H({\bf B} {\bf B}^H)^{-1} {\bf B}$, ${\bf B}$ is a known $M \times N$ matrix, with $M < N$ and $...
user avatar
  • 1
1 vote
0 answers
75 views

Kernel of the adjoint of the infinitesimal generator of Levy SDE

Consider S.D.Es driven by a combination of Brownian and non-Brownian Levy noise (like say Gamma). Then we know that the flow of the density of the S.D.E variable is given by the adjoint of the ...
user avatar
  • 2,016
1 vote
2 answers
316 views

When does the predictable $\sigma$-algebra $\mathcal{P}$ coincide with the optional $\sigma$-algebra $\mathcal{O}$?

The setup of my question is the following: Suppose that we have a measurable space $(\Omega,\mathcal{F})$ and a filtration $\mathbf{F} = (\mathcal{F}_t)_{t \geq 0}$ on it. Let $\mathcal{P}(\mathbf{F})$...
user avatar
  • 267
2 votes
1 answer
117 views

Continuous version of conditional probability distributions $( \mathcal{L}(X_t | \mathcal{G}) )_{t \geq 0}$ if $(X_t)_{t \geq 0}$ is continuous?

Let me first explain the setup: Let $(X_t)_{t \geq 0}$ be a stochastic process on some probability space $(\Omega,\mathcal{F},P)$ with values in a complete and separable metric space $E$ (e.g. $E = \...
user avatar
  • 267
2 votes
1 answer
116 views

Kalman filter distribution of observation process

Let $(X_t,Y_t)$ be a pair of stochastic processes such that $$ \begin{aligned} dX_t =& A_t X_t dt + C_t dW_t,\\ dY_t = & H_t X_t dt + K_tdB_t \end{aligned} $$ for some non-random matrix-valued ...
user avatar
  • 4,997
2 votes
0 answers
267 views

Extended Kalman Filter and its State Transition Matrix

Sorry for what might be a long post, I want to give background. Initially I had regular Kalman filter, and the state model was defined by Newtonian kinematics, with initial position 0 and speed of 2. ...
user avatar
2 votes
1 answer
308 views

Filtration exercise

I am struggling with 1.7 exercise from the Karatzas, Shreve "Brownian motion and stoch. calulus". Denote by $\mathcal{F}^X_{t_0}$ the natural filtration corresponding to a process $X:[0,\infty)\times ...
user avatar
  • 23
2 votes
0 answers
45 views

Ergodicity of differentiated processes

Let $S$ be a vector space, and $X$ a jointly-measurable random process/field with two parameters: $$ X: [0,\infty)\times\mathbb{R}\times\Omega\to S,$$ i.e. $X_{t,\theta}:\Omega\to S$ are random ...
user avatar
  • 1,655
5 votes
0 answers
233 views

A conjecture in rate distortion theory and stochastic filtering

Let $(X_t)_{t\in T}$ be a stationary random process with known and fixed law $P_X$ describing a dynamic source. This source is to be encoded real-time by an encoder $e$ into an encoded message $E_t$ ...
user avatar
  • 1,655
3 votes
1 answer
423 views

Why would one work with Kushner-FKK equation over Zakai equation?

In stochastic filtering you are interested in a process called the optimal filter $\pi_t$ which is a probability measure(d stochastic process). You can consider the unnormalized version $V_t$. The ...
user avatar
3 votes
1 answer
68 views

Filtering Mixed Discrete and Continous

Suppose I have signal process $\lambda_t$ following the dynamics \begin{equation} \begin{aligned} \zeta_t&=\mu^{\zeta}(t,{\zeta}_t)dt+\sigma^{\zeta}(t,{\zeta}_t)dW^{\zeta}_t\\ \xi_t&=\mu^{\xi}(...
user avatar
  • 4,997