Optional stopping theorem (fair game theorem) #
THIS FILE IS SYNCHRONIZED WITH MATHLIB4. Any changes to this file require a corresponding PR to mathlib4.
The optional stopping theorem states that an adapted integrable process f is a submartingale if
and only if for all bounded stopping times τ and π such that τ ≤ π, the
stopped value of f at τ has expectation smaller than its stopped value at π.
This file also contains Doob's maximal inequality: given a non-negative submartingale f, for all
ε : ℝ≥0, we have ε • μ {ε ≤ f* n} ≤ ∫ ω in {ε ≤ f* n}, f n where f* n ω = max_{k ≤ n}, f k ω.
Main results #
measure_theory.submartingale_iff_expected_stopped_value_mono: the optional stopping theorem.measure_theory.submartingale.stopped_process: the stopped process of a submartingale with respect to a stopping time is a submartingale.measure_theory.maximal_ineq: Doob's maximal inequality.
Given a submartingale f and bounded stopping times τ and π such that τ ≤ π, the
expectation of stopped_value f τ is less than or equal to the expectation of stopped_value f π.
This is the forward direction of the optional stopping theorem.
The converse direction of the optional stopping theorem, i.e. an adapted integrable process f
is a submartingale if for all bounded stopping times τ and π such that τ ≤ π, the
stopped value of f at τ has expectation smaller than its stopped value at π.
The optional stopping theorem (fair game theorem): an adapted integrable process f
is a submartingale if and only if for all bounded stopping times τ and π such that τ ≤ π, the
stopped value of f at τ has expectation smaller than its stopped value at π.
The stopped process of a submartingale with respect to a stopping time is a submartingale.
Doob's maximal inequality: Given a non-negative submartingale f, for all ε : ℝ≥0,
we have ε • μ {ε ≤ f* n} ≤ ∫ ω in {ε ≤ f* n}, f n where f* n ω = max_{k ≤ n}, f k ω.
In some literature, the Doob's maximal inequality refers to what we call Doob's Lp inequality (which is a corollary of this lemma and will be proved in an upcomming PR).