English
For any ℕ-indexed process f taking values in a normed space, the predictable part is well-defined as the sum of conditional expectations of increments with respect to the filtration.
Русский
Для любого процесса с индексами по Нат: предсказуемая часть определяется как сумма условных ожиданий приращений относительно фильтрации.
LaTeX
$$$$\mathrm{predictablePart}(f, \mathcal{F}, \mu)(n, \omega) = \sum_{i=0}^{n-1} \mu[ f(i+1) - f(i) \mid \mathcal{F}_i](\omega).$$$$
Lean4
theorem predictablePart_process_ae_eq (ℱ : Filtration ℕ m0) (μ : Measure Ω) (s : ℕ → Set Ω) (n : ℕ) :
predictablePart (process s) ℱ μ n = ∑ k ∈ Finset.range n, μ[(s (k + 1)).indicator (1 : Ω → ℝ)|ℱ k] :=
by
have := martingalePart_process_ae_eq ℱ μ s n
simp_rw [martingalePart, process, Finset.sum_sub_distrib] at this
exact sub_right_injective this