E13 Finite state Feynman Kac transfer
On this page · Permalink · Source line 4340
How to read the result
The potential below is real. Its exponential is a potential weight in general, and a survival weight when the potential is nonnegative. This distinction does not change the finite transfer formula.
Let \(K\) be column stochastic. Let \(V\) be a real potential on the states and let
For functions written as columns, the backward Markov operator is \(K^T\).
Theorem E13.1: Discrete Feynman Kac formula
Define
For terminal data \(f\),
Read proof · Read statement and proof side by side · Permalink · Source line 4351 · Claim record · Lean coverage · Code · Worked examples
Proof of Theorem E13.1
For one step,
Iterating expands the matrix product over all paths. The product of transition entries is the path probability, and the diagonal factors give the exponential potential weight.
Back to statement · Read statement and proof side by side · Permalink · Source line 4369 · Claim record · Lean coverage · Code · Worked examples
Theorem E13.2: Continuous time finite state formula
Let \(Q\) be a finite Markov generator and let \((X_t)_{t\geq0}\) be the corresponding chain. Then
If \(V\geq v_0\), then
Read proof · Read statement and proof side by side · Permalink · Source line 4379 · Claim record · Lean coverage · Code · Worked examples
Proof of Theorem E13.2
Conditioning over the first small time interval gives the backward equation
The finite system has a unique solution, namely the displayed matrix exponential. The expectation bound follows from
This finite argument is the matrix form of the Feynman Kac principle originating in [Kac1949].
Back to statement · Read statement and proof side by side · Permalink · Source line 4401 · Claim record · Lean coverage · Code · Worked examples
Theorem E13.3: Alternating propagation and killing
For every \(t\geq0\),
in operator norm. Each finite product alternates one Markov propagation step with one diagonal survival weight.
Read proof · Read statement and proof side by side · Permalink · Source line 4417 · Claim record · Lean coverage · Code · Worked examples
Proof of Theorem E13.3
Set \(B=-D_V\) and \(C=Q^T\). For bounded matrices,
uniformly in operator norm. The standard telescoping comparison with \( (I+t(B+C)/n)^n \) shows that the product converges to \({\mathrm e}^{t(B+C)}\). This is the displayed factor order, so the limit follows.
Back to statement · Read statement and proof side by side · Permalink · Source line 4430 · Claim record · Lean coverage · Code · Worked examples
Corollary E13.4: Positivity and composition
The operators
form a positive semigroup. If \(f\geq0\), then \(F_V(t)f\geq0\).
Read proof · Read statement and proof side by side · Permalink · Source line 4444 · Claim record · Lean coverage · Code · Worked examples
Proof of Corollary E13.4
Both claims follow from the expectation formula.
Back to statement · Read statement and proof side by side · Permalink · Source line 4452 · Claim record · Lean coverage · Code · Worked examples