Preliminaries. Write λ [ 0 , T ] \lambda_{[0,T]} λ [ 0 , T ] for the restricted Lebesgue measure on [ 0 , T ] [0,T] [ 0 , T ] (claim 1 of Restricted Lebesgue Measure and Integral Toolkit on a Compact Interval ); it is finite. Integrals of nonnegative functions are handled with claim 1 of Linearity and Monotonicity of the Lebesgue Integral (additivity, positive homogeneity, monotonicity) and integrable functions with claim 2 there; bounded measurable functions on a finite measure space are integrable. Sums, products and pointwise limits of measurable real functions are measurable by Arithmetic, Absolute Values, and Pointwise Limits of Measurable Real-Valued Functions , continuous functions on compact intervals by claim 4 of Measurability of Countable Suprema, Bounded Pointwise Limits, Monotone Functions, and Continuous Functions , and compositions with sequentially continuous maps (such as exp \exp exp and finite products) by Sequentially Continuous Functions of Measurable Euclidean Maps are Measurable . By Basic Properties of the Exponential Function , exp \exp exp is positive, increasing, and exp ( 0 ) = 1 \exp(0)=1 exp ( 0 ) = 1 , so exp ( − x ) ≤ 1 \exp(-x)\le1 exp ( − x ) ≤ 1 for x ≥ 0 x\ge0 x ≥ 0 ; and exp ( u ) = ∑ m ≥ 0 u m / m ! \exp(u)=\sum_{m\ge0}u^{m}/m! exp ( u ) = ∑ m ≥ 0 u m / m ! by The Real Exponential Function , the series converging absolutely. Indicators 1 [ 0 , s ] \mathbf{1}_{[0,s]} 1 [ 0 , s ] , 1 ( a , b ] \mathbf{1}_{(a,b]} 1 ( a , b ] are measurable (their sets are Borel), and 1 ( a , b ] = 1 [ 0 , b ] − 1 [ 0 , a ] \mathbf{1}_{(a,b]}=\mathbf{1}_{[0,b]}-\mathbf{1}_{[0,a]} 1 ( a , b ] = 1 [ 0 , b ] − 1 [ 0 , a ] for 0 ≤ a ≤ b ≤ T 0\le a\le b\le T 0 ≤ a ≤ b ≤ T .
Step 1: claim 1. The function μ \mu μ is measurable and bounded, hence integrable over [ 0 , T ] [0,T] [ 0 , T ] , so part (i) of Integration by Parts for Indefinite Lebesgue Integrals on a Compact Interval with f = μ f=\mu f = μ , u 0 = 0 u_0=0 u 0 = 0 shows that s ↦ M s = ∫ [ 0 , s ] μ ( u ) d u s\mapsto M_s=\int_{[0,s]}\mu(u)\,du s ↦ M s = ∫ [ 0 , s ] μ ( u ) d u is continuous on [ 0 , T ] [0,T] [ 0 , T ] ; also 0 ≤ M s ≤ μ ˉ T 0\le M_s\le\bar\mu T 0 ≤ M s ≤ μ ˉ T by monotonicity, and M M M is measurable on [ 0 , T ] [0,T] [ 0 , T ] as a continuous function.
(a) Powers of M M M . We show by induction on m ≥ 0 m\ge0 m ≥ 0 that for every s ∈ [ 0 , T ] s\in[0,T] s ∈ [ 0 , T ] ,
M s m + 1 = ( m + 1 ) ∫ [ 0 , s ] μ ( u ) M u m d u . (1.1) M_s^{m+1}=(m+1)\int_{[0,s]}\mu(u)M_u^{m}\,du .\tag{1.1} M s m + 1 = ( m + 1 ) ∫ [ 0 , s ] μ ( u ) M u m d u . ( 1.1 )
For m = 0 m=0 m = 0 this is the definition of M s M_s M s . Let m ≥ 1 m\ge1 m ≥ 1 and assume (1.1) for m − 1 m-1 m − 1 , i.e. M t m = m ∫ [ 0 , t ] μ M m − 1 M_t^{m}=m\int_{[0,t]}\mu M^{m-1} M t m = m ∫ [ 0 , t ] μ M m − 1 for every t ∈ [ 0 , T ] t\in[0,T] t ∈ [ 0 , T ] . Fix s ∈ ( 0 , T ] s\in(0,T] s ∈ ( 0 , T ] and apply part (ii) of Integration by Parts for Indefinite Lebesgue Integrals on a Compact Interval on the interval [ 0 , s ] [0,s] [ 0 , s ] (the lemma holds for every positive horizon; the integrals ∫ [ 0 , t ] \int_{[0,t]} ∫ [ 0 , t ] for t ≤ s t\le s t ≤ s are the same whether formed on [ 0 , s ] [0,s] [ 0 , s ] or on [ 0 , T ] [0,T] [ 0 , T ] , both being ∫ R \int_{\mathbb{R}} ∫ R of the same zero extension by claim 2 of Restricted Lebesgue Measure and Integral Toolkit on a Compact Interval ) with f = μ f=\mu f = μ , u 0 = 0 u_0=0 u 0 = 0 , so that u t = M t u_t=M_t u t = M t , and with g = m μ M m − 1 g=m\mu M^{m-1} g = m μ M m − 1 , v 0 = 0 v_0=0 v 0 = 0 , so that v t = M t m v_t=M_t^{m} v t = M t m by the induction hypothesis; g g g is measurable and bounded by m μ ˉ ( μ ˉ T ) m − 1 m\bar\mu(\bar\mu T)^{m-1} m μ ˉ ( μ ˉ T ) m − 1 , hence integrable. The conclusion reads
M s M s m = ∫ [ 0 , s ] ( μ ( u ) M u m + M u m μ ( u ) M u m − 1 ) d u = ( m + 1 ) ∫ [ 0 , s ] μ ( u ) M u m d u , M_s\,M_s^{m}=\int_{[0,s]}\bigl(\mu(u)M_u^{m}+M_u\,m\,\mu(u)M_u^{m-1}\bigr)\,du=(m+1)\int_{[0,s]}\mu(u)M_u^{m}\,du , M s M s m = ∫ [ 0 , s ] ( μ ( u ) M u m + M u m μ ( u ) M u m − 1 ) d u = ( m + 1 ) ∫ [ 0 , s ] μ ( u ) M u m d u ,
which is (1.1) for m m m at s s s ; at s = 0 s=0 s = 0 both sides of (1.1) vanish. Consequently, for 0 ≤ a ≤ b ≤ T 0\le a\le b\le T 0 ≤ a ≤ b ≤ T , by 1 ( a , b ] = 1 [ 0 , b ] − 1 [ 0 , a ] \mathbf{1}_{(a,b]}=\mathbf{1}_{[0,b]}-\mathbf{1}_{[0,a]} 1 ( a , b ] = 1 [ 0 , b ] − 1 [ 0 , a ] and linearity,
∫ ( a , b ] μ ( u ) M u m d u = M b m + 1 − M a m + 1 m + 1 . (1.2) \int_{(a,b]}\mu(u)M_u^{m}\,du=\frac{M_b^{m+1}-M_a^{m+1}}{m+1}.\tag{1.2} ∫ ( a , b ] μ ( u ) M u m d u = m + 1 M b m + 1 − M a m + 1 . ( 1.2 )
(b) The series. For u ∈ [ 0 , T ] u\in[0,T] u ∈ [ 0 , T ] put S n ( u ) = ∑ m = 0 n ( − 1 ) m m ! μ ( u ) M u m S_n(u)=\sum_{m=0}^{n}\frac{(-1)^{m}}{m!}\mu(u)M_u^{m} S n ( u ) = ∑ m = 0 n m ! ( − 1 ) m μ ( u ) M u m . Each S n S_n S n is measurable, S n ( u ) → μ ( u ) exp ( − M u ) S_n(u)\to\mu(u)\exp(-M_u) S n ( u ) → μ ( u ) exp ( − M u ) as n → ∞ n\to\infty n → ∞ by the series for exp ( − M u ) \exp(-M_u) exp ( − M u ) , and ∣ S n ( u ) ∣ ≤ μ ˉ ∑ m ≥ 0 ( μ ˉ T ) m / m ! = μ ˉ exp ( μ ˉ T ) |S_n(u)|\le\bar\mu\sum_{m\ge0}(\bar\mu T)^{m}/m!=\bar\mu\exp(\bar\mu T) ∣ S n ( u ) ∣ ≤ μ ˉ ∑ m ≥ 0 ( μ ˉ T ) m / m ! = μ ˉ exp ( μ ˉ T ) , a constant, integrable on the finite measure space ( [ 0 , T ] , λ [ 0 , T ] ) ([0,T],\lambda_{[0,T]}) ([ 0 , T ] , λ [ 0 , T ] ) . In particular u ↦ μ ( u ) exp ( − M u ) u\mapsto\mu(u)\exp(-M_u) u ↦ μ ( u ) exp ( − M u ) is measurable (a pointwise limit), and by Dominated Convergence Theorem and (1.2),
∫ ( a , b ] μ ( u ) exp ( − M u ) d u = lim n → ∞ ∫ ( a , b ] S n ( u ) d u = lim n → ∞ ∑ m = 0 n ( − 1 ) m ( m + 1 ) ! ( M b m + 1 − M a m + 1 ) . \int_{(a,b]}\mu(u)\exp(-M_u)\,du=\lim_{n\to\infty}\int_{(a,b]}S_n(u)\,du=\lim_{n\to\infty}\sum_{m=0}^{n}\frac{(-1)^{m}}{(m+1)!}\bigl(M_b^{m+1}-M_a^{m+1}\bigr). ∫ ( a , b ] μ ( u ) exp ( − M u ) d u = n → ∞ lim ∫ ( a , b ] S n ( u ) d u = n → ∞ lim m = 0 ∑ n ( m + 1 )! ( − 1 ) m ( M b m + 1 − M a m + 1 ) .
For real x x x , ∑ m = 0 n ( − 1 ) m x m + 1 ( m + 1 ) ! = − ∑ j = 1 n + 1 ( − x ) j j ! → − ( exp ( − x ) − 1 ) = 1 − exp ( − x ) \sum_{m=0}^{n}\frac{(-1)^{m}x^{m+1}}{(m+1)!}=-\sum_{j=1}^{n+1}\frac{(-x)^{j}}{j!}\to-(\exp(-x)-1)=1-\exp(-x) ∑ m = 0 n ( m + 1 )! ( − 1 ) m x m + 1 = − ∑ j = 1 n + 1 j ! ( − x ) j → − ( exp ( − x ) − 1 ) = 1 − exp ( − x ) by the series for exp ( − x ) \exp(-x) exp ( − x ) . Applying this with x = M b x=M_b x = M b and x = M a x=M_a x = M a and subtracting (limits of sums, Arithmetic of Limits of Real Sequences ) gives ∫ ( a , b ] μ ( u ) exp ( − M u ) d u = ( 1 − exp ( − M b ) ) − ( 1 − exp ( − M a ) ) = exp ( − M a ) − exp ( − M b ) \int_{(a,b]}\mu(u)\exp(-M_u)\,du=(1-\exp(-M_b))-(1-\exp(-M_a))=\exp(-M_a)-\exp(-M_b) ∫ ( a , b ] μ ( u ) exp ( − M u ) d u = ( 1 − exp ( − M b )) − ( 1 − exp ( − M a )) = exp ( − M a ) − exp ( − M b ) , which is claim 1.
Step 2: claim 2. Cells and transports. For k ≥ 1 k\ge1 k ≥ 1 and v ∈ V k v\in V^k v ∈ V k let χ k , v : D k ( T ) → C k , v \chi_{k,v}:D_k(T)\to C_{k,v} χ k , v : D k ( T ) → C k , v , χ k , v ( t ) = ( k , t , v ) \chi_{k,v}(t)=(k,t,v) χ k , v ( t ) = ( k , t , v ) , be the bijection along which the cell C k , v C_{k,v} C k , v carries the transport of the restriction of ( R k , B k , λ k ) (\mathbb{R}^k,\mathcal{B}_k,\lambda_k) ( R k , B k , λ k ) to D k ( T ) D_k(T) D k ( T ) (The Observation Record Space , with B k \mathcal{B}_k B k and λ k \lambda_k λ k as in Finite Products of Lebesgue Measure and Coordinate Integration on R l \mathbb{R}^l R l ); write D k \mathcal{D}_k D k for the σ \sigma σ -algebra of that restriction (the Borel subsets of D k ( T ) D_k(T) D k ( T ) , claim 1 of Assembly of Measure Spaces: Restriction, Transport, One-Point Spaces, and Countable Disjoint Unions ) and λ D k \lambda_{D_k} λ D k for its measure, which is finite with λ D k ( D k ( T ) ) = T k / k ! \lambda_{D_k}(D_k(T))=T^k/k! λ D k ( D k ( T )) = T k / k ! by The Ordered Time Simplex: Borel Measurability and Volume . By claims 2, 3 and 4 of Assembly of Measure Spaces: Restriction, Transport, One-Point Spaces, and Countable Disjoint Unions , a function h : R → [ 0 , ∞ ] h:\mathbf{R}\to[0,\infty] h : R → [ 0 , ∞ ] is R \mathcal{R} R -measurable if and only if h ∘ χ k , v h\circ\chi_{k,v} h ∘ χ k , v is D k \mathcal{D}_k D k -measurable for every k ≥ 1 k\ge1 k ≥ 1 and v ∈ V k v\in V^k v ∈ V k (there is no condition on the one-point cell C ∅ C_\emptyset C ∅ ), and then, by claim 4(c) there and the change of variables of claim 2 of Image Measures, Measures with Densities, and Change of Variables along the transport,
∫ R h d ρ = h ( r ∅ ) + ∑ k ≥ 1 ∑ v ∈ V k ∫ D k ( T ) h ∘ χ k , v d λ D k , ∫ R ( k ) h d ρ = ∑ v ∈ V k ∫ D k ( T ) h ∘ χ k , v d λ D k ( k ≥ 1 ) , ∫ R ( 0 ) h d ρ = h ( r ∅ ) , (2.1) \int_{\mathbf{R}}h\,d\rho=h(r_\emptyset)+\sum_{k\ge1}\sum_{v\in V^k}\int_{D_k(T)}h\circ\chi_{k,v}\,d\lambda_{D_k},\qquad \int_{\mathbf{R}^{(k)}}h\,d\rho=\sum_{v\in V^k}\int_{D_k(T)}h\circ\chi_{k,v}\,d\lambda_{D_k}\ (k\ge1),\qquad\int_{\mathbf{R}^{(0)}}h\,d\rho=h(r_\emptyset),\tag{2.1} ∫ R h d ρ = h ( r ∅ ) + k ≥ 1 ∑ v ∈ V k ∑ ∫ D k ( T ) h ∘ χ k , v d λ D k , ∫ R ( k ) h d ρ = v ∈ V k ∑ ∫ D k ( T ) h ∘ χ k , v d λ D k ( k ≥ 1 ) , ∫ R ( 0 ) h d ρ = h ( r ∅ ) , ( 2.1 )
the first sum being over the countable cell family in the finite-partial-sum sense of that lemma (the second and third identities follow from the first applied to h 1 R ( k ) h\mathbf{1}_{\mathbf{R}^{(k)}} h 1 R ( k ) , whose restrictions to the other cells vanish). Moreover, by claim 1 of Assembly of Measure Spaces: Restriction, Transport, One-Point Spaces, and Countable Disjoint Unions , ∫ D k ( T ) h ∘ χ k , v d λ D k = ∫ R k h ∘ χ k , v ~ d λ k \int_{D_k(T)}h\circ\chi_{k,v}\,d\lambda_{D_k}=\int_{\mathbb{R}^k}\widetilde{h\circ\chi_{k,v}}\,d\lambda_k ∫ D k ( T ) h ∘ χ k , v d λ D k = ∫ R k h ∘ χ k , v d λ k with ~ \widetilde{\ } the zero extension.
Measurability on a cell. Fix k ≥ 1 k\ge1 k ≥ 1 , v ∈ V k v\in V^k v ∈ V k , write χ = χ k , v \chi=\chi_{k,v} χ = χ k , v , and let Ξ : [ 0 , T ] × D k ( T ) → [ 0 , T ] × R \Xi:[0,T]\times D_k(T)\to[0,T]\times\mathbf{R} Ξ : [ 0 , T ] × D k ( T ) → [ 0 , T ] × R , Ξ ( s , t ) = ( s , χ ( t ) ) \Xi(s,t)=(s,\chi(t)) Ξ ( s , t ) = ( s , χ ( t )) . The trace on [ 0 , T ] × D k ( T ) [0,T]\times D_k(T) [ 0 , T ] × D k ( T ) of every set in B 1 + k \mathcal{B}_{1+k} B 1 + k belongs to the product σ \sigma σ -algebra B [ 0 , T ] ⊗ D k \mathcal{B}_{[0,T]}\otimes\mathcal{D}_k B [ 0 , T ] ⊗ D k : the trace of a Borel rectangle A 1 × A 2 × ⋯ × A 1 + k A_1\times A_2\times\dots\times A_{1+k} A 1 × A 2 × ⋯ × A 1 + k is ( A 1 ∩ [ 0 , T ] ) × ( ( A 2 × ⋯ × A 1 + k ) ∩ D k ( T ) ) (A_1\cap[0,T])\times\bigl((A_2\times\dots\times A_{1+k})\cap D_k(T)\bigr) ( A 1 ∩ [ 0 , T ]) × ( ( A 2 × ⋯ × A 1 + k ) ∩ D k ( T ) ) , a measurable rectangle (the second factor being the trace on D k ( T ) D_k(T) D k ( T ) of a member of B k \mathcal{B}_k B k ), and the sets in B 1 + k \mathcal{B}_{1+k} B 1 + k whose trace lies in the product σ \sigma σ -algebra form a σ \sigma σ -algebra containing the Borel rectangles, which generate B 1 + k \mathcal{B}_{1+k} B 1 + k (claim 1 of Finite Products of Lebesgue Measure and Coordinate Integration on R l \mathbb{R}^l R l ). The map Ξ \Xi Ξ is measurable with respect to B [ 0 , T ] ⊗ D k \mathcal{B}_{[0,T]}\otimes\mathcal{D}_k B [ 0 , T ] ⊗ D k and B [ 0 , T ] ⊗ R \mathcal{B}_{[0,T]}\otimes\mathcal{R} B [ 0 , T ] ⊗ R : the preimage of a rectangle A × B A\times B A × B (A ∈ B [ 0 , T ] A\in\mathcal{B}_{[0,T]} A ∈ B [ 0 , T ] , B ∈ R B\in\mathcal{R} B ∈ R ) is A × χ − 1 ( B ∩ C k , v ) A\times\chi^{-1}(B\cap C_{k,v}) A × χ − 1 ( B ∩ C k , v ) , and B ∩ C k , v B\cap C_{k,v} B ∩ C k , v belongs to the cell σ \sigma σ -algebra (claim 4(a) of Assembly of Measure Spaces: Restriction, Transport, One-Point Spaces, and Countable Disjoint Unions ), whose members are the images χ ( E ) \chi(E) χ ( E ) , E ∈ D k E\in\mathcal{D}_k E ∈ D k (claim 2 there); rectangles generate the product σ \sigma σ -algebra, so the generator criterion of Measurable Function and Real-Valued Measurable Function applies. Hence, for every υ ∈ V \upsilon\in V υ ∈ V , ( s , t ) ↦ λ s υ ( χ ( t ) ) = λ υ ∘ Ξ ( s , t ) (s,t)\mapsto\lambda^\upsilon_s(\chi(t))=\lambda^\upsilon\circ\Xi(s,t) ( s , t ) ↦ λ s υ ( χ ( t )) = λ υ ∘ Ξ ( s , t ) is B [ 0 , T ] ⊗ D k \mathcal{B}_{[0,T]}\otimes\mathcal{D}_k B [ 0 , T ] ⊗ D k -measurable by condition (i) of Causal Intensity on the Observation Record Space , and so is ( s , t ) ↦ λ s t o t ( χ ( t ) ) (s,t)\mapsto\lambda^{\mathrm{tot}}_s(\chi(t)) ( s , t ) ↦ λ s tot ( χ ( t )) . Consequently: (a) for each i ≤ k i\le k i ≤ k , t ↦ λ t i v i ( χ ( t ) ) t\mapsto\lambda^{v_i}_{t_i}(\chi(t)) t ↦ λ t i v i ( χ ( t )) is D k \mathcal{D}_k D k -measurable, being the composition of the previous map with t ↦ ( t i , t ) t\mapsto(t_i,t) t ↦ ( t i , t ) , which is measurable from D k \mathcal{D}_k D k to B [ 0 , T ] ⊗ D k \mathcal{B}_{[0,T]}\otimes\mathcal{D}_k B [ 0 , T ] ⊗ D k by the generator criterion of Measurable Function and Real-Valued Measurable Function , the preimage of a rectangle A × B A\times B A × B being { t ∈ B : t i ∈ A } ∈ D k \{t\in B:t_i\in A\}\in\mathcal{D}_k { t ∈ B : t i ∈ A } ∈ D k (the coordinate projection t ↦ t i t\mapsto t_i t ↦ t i is measurable by claim 1 of The Borel Sigma-Algebra of a Euclidean Space as a Product, and Measurability of Projections, Sequentially Continuous Maps, and Open and Closed Sets ); (b) the function ( s , t ) ↦ 1 { s ≤ t k } λ s t o t ( χ ( t ) ) (s,t)\mapsto\mathbf{1}\{s\le t_k\}\lambda^{\mathrm{tot}}_s(\chi(t)) ( s , t ) ↦ 1 { s ≤ t k } λ s tot ( χ ( t )) is jointly measurable (the set { ( s , t ) : s ≤ t k } \{(s,t):s\le t_k\} {( s , t ) : s ≤ t k } is closed in R 1 + k \mathbb{R}^{1+k} R 1 + k , hence Borel by claim 4 of The Borel Sigma-Algebra of a Euclidean Space as a Product, and Measurability of Projections, Sequentially Continuous Maps, and Open and Closed Sets ), so by the Tonelli theorem (Tonelli and Fubini Theorems , for the finite measures λ [ 0 , T ] \lambda_{[0,T]} λ [ 0 , T ] and λ D k \lambda_{D_k} λ D k ) the map t ↦ Λ t k ( χ ( t ) ) = ∫ [ 0 , T ] 1 { s ≤ t k } λ s t o t ( χ ( t ) ) d s t\mapsto\Lambda_{t_k}(\chi(t))=\int_{[0,T]}\mathbf{1}\{s\le t_k\}\lambda^{\mathrm{tot}}_s(\chi(t))\,ds t ↦ Λ t k ( χ ( t )) = ∫ [ 0 , T ] 1 { s ≤ t k } λ s tot ( χ ( t )) d s is D k \mathcal{D}_k D k -measurable, and likewise t ↦ Λ s ( χ ( t ) ) t\mapsto\Lambda_s(\chi(t)) t ↦ Λ s ( χ ( t )) for every fixed s ∈ [ 0 , T ] s\in[0,T] s ∈ [ 0 , T ] (with the Borel set { ( u , t ) : u ≤ s } \{(u,t):u\le s\} {( u , t ) : u ≤ s } in place of { ( u , t ) : u ≤ t k } \{(u,t):u\le t_k\} {( u , t ) : u ≤ t k } ), in particular t ↦ Λ T ( χ ( t ) ) t\mapsto\Lambda_T(\chi(t)) t ↦ Λ T ( χ ( t )) ; as Λ s ( r ∅ ) \Lambda_s(r_\emptyset) Λ s ( r ∅ ) imposes no condition, r ↦ Λ s ( r ) r\mapsto\Lambda_s(r) r ↦ Λ s ( r ) is R \mathcal{R} R -measurable for every s s s . Therefore q ∘ χ = ( ∏ i ≤ k λ t i v i ∘ χ ) exp ( − Λ t k ∘ χ ) q\circ\chi=\bigl(\prod_{i\le k}\lambda^{v_i}_{t_i}\circ\chi\bigr)\exp(-\Lambda_{t_k}\circ\chi) q ∘ χ = ( ∏ i ≤ k λ t i v i ∘ χ ) exp ( − Λ t k ∘ χ ) and ℓ λ ∘ χ = ( ∏ i ≤ k λ t i v i ∘ χ ) exp ( − Λ T ∘ χ ) \ell_\lambda\circ\chi=\bigl(\prod_{i\le k}\lambda^{v_i}_{t_i}\circ\chi\bigr)\exp(-\Lambda_T\circ\chi) ℓ λ ∘ χ = ( ∏ i ≤ k λ t i v i ∘ χ ) exp ( − Λ T ∘ χ ) are D k \mathcal{D}_k D k -measurable, and q q q , ℓ λ \ell_\lambda ℓ λ are R \mathcal{R} R -measurable.
Bounds. For r = ( k , t , v ) ∈ R ( k ) r=(k,t,v)\in\mathbf{R}^{(k)} r = ( k , t , v ) ∈ R ( k ) : each factor λ t i v i ( r ) \lambda^{v_i}_{t_i}(r) λ t i v i ( r ) lies in [ 0 , λ ˉ ] [0,\bar\lambda] [ 0 , λ ˉ ] and exp ( − Λ t k ( r ) ) ≤ 1 \exp(-\Lambda_{t_k}(r))\le1 exp ( − Λ t k ( r )) ≤ 1 , so 0 ≤ q ( r ) ≤ λ ˉ k 0\le q(r)\le\bar\lambda^k 0 ≤ q ( r ) ≤ λ ˉ k . Since 1 [ 0 , T ] = 1 [ 0 , t k ] + 1 ( t k , T ] \mathbf{1}_{[0,T]}=\mathbf{1}_{[0,t_k]}+\mathbf{1}_{(t_k,T]} 1 [ 0 , T ] = 1 [ 0 , t k ] + 1 ( t k , T ] and λ t o t ≥ 0 \lambda^{\mathrm{tot}}\ge0 λ tot ≥ 0 , linearity gives Λ T ( r ) = Λ t k ( r ) + ∫ ( t k , T ] λ u t o t ( r ) d u ≥ Λ t k ( r ) \Lambda_T(r)=\Lambda_{t_k}(r)+\int_{(t_k,T]}\lambda^{\mathrm{tot}}_u(r)\,du\ge\Lambda_{t_k}(r) Λ T ( r ) = Λ t k ( r ) + ∫ ( t k , T ] λ u tot ( r ) d u ≥ Λ t k ( r ) , and the functional equation of exp \exp exp (claim 1 of Basic Properties of the Exponential Function ) gives ℓ λ ( r ) = q ( r ) exp ( − ( Λ T ( r ) − Λ t k ( r ) ) ) ≤ q ( r ) \ell_\lambda(r)=q(r)\exp(-(\Lambda_T(r)-\Lambda_{t_k}(r)))\le q(r) ℓ λ ( r ) = q ( r ) exp ( − ( Λ T ( r ) − Λ t k ( r ))) ≤ q ( r ) . Finally, by (2.1) and monotonicity, Q n ≤ ∑ v ∈ V n λ ˉ n λ D n ( D n ( T ) ) = l ~ n λ ˉ n T n / n ! Q_n\le\sum_{v\in V^n}\bar\lambda^n\lambda_{D_n}(D_n(T))=\tilde{l}^n\bar\lambda^nT^n/n! Q n ≤ ∑ v ∈ V n λ ˉ n λ D n ( D n ( T )) = l ~ n λ ˉ n T n / n ! for n ≥ 1 n\ge1 n ≥ 1 (V n V^n V n having l ~ n \tilde{l}^n l ~ n members), and Q 0 = q ( r ∅ ) = 1 Q_0=q(r_\emptyset)=1 Q 0 = q ( r ∅ ) = 1 .
Step 3: claim 3. Fix n ≥ 0 n\ge0 n ≥ 0 and r = ( n , t , v ) ∈ R ( n ) r=(n,t,v)\in\mathbf{R}^{(n)} r = ( n , t , v ) ∈ R ( n ) (with t 0 = 0 t_0=0 t 0 = 0 and empty tuples when n = 0 n=0 n = 0 ). For s ∈ ( t n , T ] s\in(t_n,T] s ∈ ( t n , T ] and v ′ ∈ V v'\in V v ′ ∈ V define the appended record
r\sqcup(s,v')=\bigl(n+1,\ (t_1,\dots,t_n,s),\ (v_1,\dots,v_n,v')\bigr)\in\mathbf{R}^{(n)}+1)},
a record since 0 < t 1 < ⋯ < t n < s ≤ T 0<t_1<\dots<t_n<s\le T 0 < t 1 < ⋯ < t n < s ≤ T . Causality. For u ∈ [ 0 , s ] u\in[0,s] u ∈ [ 0 , s ] the strict prefix maps satisfy π u − ( r ⊔ ( s , v ′ ) ) = π u − ( r ) \pi_{u-}(r\sqcup(s,v'))=\pi_{u-}(r) π u − ( r ⊔ ( s , v ′ )) = π u − ( r ) : both keep exactly the events of r r r with time < u <u < u , the appended time s ≥ u s\ge u s ≥ u not being < u <u < u . Condition (ii) of Causal Intensity on the Observation Record Space therefore gives λ u υ ( r ⊔ ( s , v ′ ) ) = λ u υ ( r ) \lambda^\upsilon_u(r\sqcup(s,v'))=\lambda^\upsilon_u(r) λ u υ ( r ⊔ ( s , v ′ )) = λ u υ ( r ) for all u ∈ [ 0 , s ] u\in[0,s] u ∈ [ 0 , s ] and υ ∈ V \upsilon\in V υ ∈ V . In particular λ t i v i ( r ⊔ ( s , v ′ ) ) = λ t i v i ( r ) \lambda^{v_i}_{t_i}(r\sqcup(s,v'))=\lambda^{v_i}_{t_i}(r) λ t i v i ( r ⊔ ( s , v ′ )) = λ t i v i ( r ) for i ≤ n i\le n i ≤ n (as t i < s t_i<s t i < s ), λ s v ′ ( r ⊔ ( s , v ′ ) ) = λ s v ′ ( r ) \lambda^{v'}_s(r\sqcup(s,v'))=\lambda^{v'}_s(r) λ s v ′ ( r ⊔ ( s , v ′ )) = λ s v ′ ( r ) , and Λ s ( r ⊔ ( s , v ′ ) ) = ∫ [ 0 , s ] λ u t o t ( r ⊔ ( s , v ′ ) ) d u = Λ s ( r ) \Lambda_s(r\sqcup(s,v'))=\int_{[0,s]}\lambda^{\mathrm{tot}}_u(r\sqcup(s,v'))\,du=\Lambda_s(r) Λ s ( r ⊔ ( s , v ′ )) = ∫ [ 0 , s ] λ u tot ( r ⊔ ( s , v ′ )) d u = Λ s ( r ) . With P ( r ) = ∏ i ≤ n λ t i v i ( r ) P(r)=\prod_{i\le n}\lambda^{v_i}_{t_i}(r) P ( r ) = ∏ i ≤ n λ t i v i ( r ) this yields
q ( r ⊔ ( s , v ′ ) ) = P ( r ) λ s v ′ ( r ) exp ( − Λ s ( r ) ) , ∑ v ′ ∈ V q ( r ⊔ ( s , v ′ ) ) = P ( r ) λ s t o t ( r ) exp ( − Λ s ( r ) ) ( s ∈ ( t n , T ] ) . q\bigl(r\sqcup(s,v')\bigr)=P(r)\,\lambda^{v'}_s(r)\exp(-\Lambda_s(r)),\qquad \sum_{v'\in V}q\bigl(r\sqcup(s,v')\bigr)=P(r)\,\lambda^{\mathrm{tot}}_s(r)\exp(-\Lambda_s(r))\qquad(s\in(t_n,T]). q ( r ⊔ ( s , v ′ ) ) = P ( r ) λ s v ′ ( r ) exp ( − Λ s ( r )) , v ′ ∈ V ∑ q ( r ⊔ ( s , v ′ ) ) = P ( r ) λ s tot ( r ) exp ( − Λ s ( r )) ( s ∈ ( t n , T ]) .
Applying claim 1 with μ = λ ⋅ t o t ( r ) \mu=\lambda^{\mathrm{tot}}_\cdot(r) μ = λ ⋅ tot ( r ) (measurable and bounded by l ~ λ ˉ \tilde{l}\bar\lambda l ~ λ ˉ , as noted in Likelihood of a Causal Intensity on the Observation Record Space ), M = Λ ⋅ ( r ) M=\Lambda_\cdot(r) M = Λ ⋅ ( r ) , a = t n a=t_n a = t n and b = T b=T b = T , and using ℓ λ ( r ) = P ( r ) exp ( − Λ T ( r ) ) \ell_\lambda(r)=P(r)\exp(-\Lambda_T(r)) ℓ λ ( r ) = P ( r ) exp ( − Λ T ( r )) and q ( r ) = P ( r ) exp ( − Λ t n ( r ) ) q(r)=P(r)\exp(-\Lambda_{t_n}(r)) q ( r ) = P ( r ) exp ( − Λ t n ( r )) ,
∑ v ′ ∈ V ∫ ( t n , T ] q ( r ⊔ ( s , v ′ ) ) d s = P ( r ) ( exp ( − Λ t n ( r ) ) − exp ( − Λ T ( r ) ) ) = q ( r ) − ℓ λ ( r ) . (3.1) \sum_{v'\in V}\int_{(t_n,T]}q\bigl(r\sqcup(s,v')\bigr)\,ds=P(r)\bigl(\exp(-\Lambda_{t_n}(r))-\exp(-\Lambda_T(r))\bigr)=q(r)-\ell_\lambda(r).\tag{3.1} v ′ ∈ V ∑ ∫ ( t n , T ] q ( r ⊔ ( s , v ′ ) ) d s = P ( r ) ( exp ( − Λ t n ( r )) − exp ( − Λ T ( r )) ) = q ( r ) − ℓ λ ( r ) . ( 3.1 )
Integration over R ( n ) \mathbf{R}^{(n)} R ( n ) . Fix v ∈ V n v\in V^n v ∈ V n and v ′ ∈ V v'\in V v ′ ∈ V and let q ~ \tilde q q ~ be the zero extension to R n + 1 \mathbb{R}^{n+1} R n + 1 of q ∘ χ n + 1 , ( v , v ′ ) q\circ\chi_{n+1,(v,v')} q ∘ χ n + 1 , ( v , v ′ ) , measurable by Step 2 and claim 1 of Assembly of Measure Spaces: Restriction, Transport, One-Point Spaces, and Countable Disjoint Unions . Under the identification R n + 1 = R n × R \mathbb{R}^{n+1}=\mathbb{R}^n\times\mathbb{R} R n + 1 = R n × R of Finite Products of Lebesgue Measure and Coordinate Integration on R l \mathbb{R}^l R l , λ n + 1 = λ n ⊗ λ \lambda_{n+1}=\lambda_n\otimes\lambda λ n + 1 = λ n ⊗ λ and 1 D n + 1 ( T ) ( t , s ) = 1 D n ( T ) ( t ) 1 ( t n , T ] ( s ) \mathbf{1}_{D_{n+1}(T)}(t,s)=\mathbf{1}_{D_n(T)}(t)\mathbf{1}_{(t_n,T]}(s) 1 D n + 1 ( T ) ( t , s ) = 1 D n ( T ) ( t ) 1 ( t n , T ] ( s ) (for n = 0 n=0 n = 0 read R 0 × R \mathbb{R}^0\times\mathbb{R} R 0 × R as R \mathbb{R} R , 1 D 0 ( T ) = 1 \mathbf{1}_{D_0(T)}=1 1 D 0 ( T ) = 1 , t 0 = 0 t_0=0 t 0 = 0 , and D 1 ( T ) = ( 0 , T ] D_1(T)=(0,T] D 1 ( T ) = ( 0 , T ] ), while q ( r ⊔ ( s , v ′ ) ) = q ~ ( t , s ) q(r\sqcup(s,v'))=\tilde q(t,s) q ( r ⊔ ( s , v ′ )) = q ~ ( t , s ) for t ∈ D n ( T ) t\in D_n(T) t ∈ D n ( T ) , s ∈ ( t n , T ] s\in(t_n,T] s ∈ ( t n , T ] . The Tonelli theorem (Tonelli and Fubini Theorems ) for λ n ⊗ λ \lambda_n\otimes\lambda λ n ⊗ λ applied to the nonnegative measurable function 1 D n + 1 ( T ) q ~ \mathbf{1}_{D_{n+1}(T)}\tilde q 1 D n + 1 ( T ) q ~ gives
∫ R n 1 D n ( T ) ( t ) ( ∫ R 1 ( t n , T ] ( s ) q ~ ( t , s ) λ ( d s ) ) λ n ( d t ) = ∫ R n + 1 1 D n + 1 ( T ) q ~ d λ n + 1 = ∫ D n + 1 ( T ) q ∘ χ n + 1 , ( v , v ′ ) d λ D n + 1 , \int_{\mathbb{R}^n}\mathbf{1}_{D_n(T)}(t)\Bigl(\int_{\mathbb{R}}\mathbf{1}_{(t_n,T]}(s)\,\tilde q(t,s)\,\lambda(ds)\Bigr)\lambda_n(dt)=\int_{\mathbb{R}^{n+1}}\mathbf{1}_{D_{n+1}(T)}\tilde q\,d\lambda_{n+1}=\int_{D_{n+1}(T)}q\circ\chi_{n+1,(v,v')}\,d\lambda_{D_{n+1}} , ∫ R n 1 D n ( T ) ( t ) ( ∫ R 1 ( t n , T ] ( s ) q ~ ( t , s ) λ ( d s ) ) λ n ( d t ) = ∫ R n + 1 1 D n + 1 ( T ) q ~ d λ n + 1 = ∫ D n + 1 ( T ) q ∘ χ n + 1 , ( v , v ′ ) d λ D n + 1 ,
the inner integral being ∫ ( t n , T ] q ( r ⊔ ( s , v ′ ) ) d s \int_{(t_n,T]}q(r\sqcup(s,v'))\,ds ∫ ( t n , T ] q ( r ⊔ ( s , v ′ )) d s for r = χ n , v ( t ) r=\chi_{n,v}(t) r = χ n , v ( t ) (claim 2 of Restricted Lebesgue Measure and Integral Toolkit on a Compact Interval , the integrand vanishing off [ 0 , T ] [0,T] [ 0 , T ] ), and t ↦ t\mapsto t ↦ (inner integral) being measurable by Tonelli. Summing over v ′ ∈ V v'\in V v ′ ∈ V and v ∈ V n v\in V^n v ∈ V n , using (2.1) for n n n and for n + 1 n+1 n + 1 (the pairs ( v , v ′ ) (v,v') ( v , v ′ ) running through V n + 1 V^{n+1} V n + 1 ), and inserting (3.1) pointwise on R ( n ) \mathbf{R}^{(n)} R ( n ) :
Q n + 1 = ∑ v ∈ V n ∫ D n ( T ) ( ∑ v ′ ∈ V ∫ ( t n , T ] q ( χ n , v ( t ) ⊔ ( s , v ′ ) ) d s ) λ D n ( d t ) = ∫ R ( n ) ( q − ℓ λ ) d ρ = Q n − ∫ R ( n ) ℓ λ d ρ , Q_{n+1}=\sum_{v\in V^n}\int_{D_n(T)}\Bigl(\sum_{v'\in V}\int_{(t_n,T]}q(\chi_{n,v}(t)\sqcup(s,v'))\,ds\Bigr)\lambda_{D_n}(dt)=\int_{\mathbf{R}^{(n)}}(q-\ell_\lambda)\,d\rho=Q_n-\int_{\mathbf{R}^{(n)}}\ell_\lambda\,d\rho , Q n + 1 = v ∈ V n ∑ ∫ D n ( T ) ( v ′ ∈ V ∑ ∫ ( t n , T ] q ( χ n , v ( t ) ⊔ ( s , v ′ )) d s ) λ D n ( d t ) = ∫ R ( n ) ( q − ℓ λ ) d ρ = Q n − ∫ R ( n ) ℓ λ d ρ ,
the last step by additivity, both q q q and ℓ λ \ell_\lambda ℓ λ being nonnegative with finite integrals over R ( n ) \mathbf{R}^{(n)} R ( n ) (claim 2) and q − ℓ λ ≥ 0 q-\ell_\lambda\ge0 q − ℓ λ ≥ 0 . For n = 0 n=0 n = 0 the same computation reads Q 1 = ∑ v ′ ∫ D 1 ( T ) q ∘ χ 1 , v ′ d λ D 1 = ∑ v ′ ∫ ( 0 , T ] q ( r ∅ ⊔ ( s , v ′ ) ) d s = q ( r ∅ ) − ℓ λ ( r ∅ ) = Q 0 − ∫ R ( 0 ) ℓ λ d ρ Q_1=\sum_{v'}\int_{D_1(T)}q\circ\chi_{1,v'}\,d\lambda_{D_1}=\sum_{v'}\int_{(0,T]}q(r_\emptyset\sqcup(s,v'))\,ds=q(r_\emptyset)-\ell_\lambda(r_\emptyset)=Q_0-\int_{\mathbf{R}^{(0)}}\ell_\lambda\,d\rho Q 1 = ∑ v ′ ∫ D 1 ( T ) q ∘ χ 1 , v ′ d λ D 1 = ∑ v ′ ∫ ( 0 , T ] q ( r ∅ ⊔ ( s , v ′ )) d s = q ( r ∅ ) − ℓ λ ( r ∅ ) = Q 0 − ∫ R ( 0 ) ℓ λ d ρ . This is claim 3.
Step 4: claim 4. By claim 3, ∫ R ( k ) ℓ λ d ρ = Q k − Q k + 1 \int_{\mathbf{R}^{(k)}}\ell_\lambda\,d\rho=Q_k-Q_{k+1} ∫ R ( k ) ℓ λ d ρ = Q k − Q k + 1 for every k ≥ 0 k\ge0 k ≥ 0 , so the sum over k ≤ n k\le n k ≤ n telescopes to Q 0 − Q n + 1 = 1 − Q n + 1 Q_0-Q_{n+1}=1-Q_{n+1} Q 0 − Q n + 1 = 1 − Q n + 1 . By (2.1), ∫ R ℓ λ d ρ \int_{\mathbf{R}}\ell_\lambda\,d\rho ∫ R ℓ λ d ρ is the least upper bound of the finite partial sums of the cell integrals of ℓ λ \ell_\lambda ℓ λ ; every finite family of cells is contained in the family of cells with at most n n n events for some n n n , whose (finite) partial sum is ∑ k ≤ n ∫ R ( k ) ℓ λ d ρ = 1 − Q n + 1 ≤ 1 \sum_{k\le n}\int_{\mathbf{R}^{(k)}}\ell_\lambda\,d\rho=1-Q_{n+1}\le1 ∑ k ≤ n ∫ R ( k ) ℓ λ d ρ = 1 − Q n + 1 ≤ 1 , and conversely each of these is a finite partial sum. Hence ∫ R ℓ λ d ρ = sup n ( 1 − Q n + 1 ) \int_{\mathbf{R}}\ell_\lambda\,d\rho=\sup_n(1-Q_{n+1}) ∫ R ℓ λ d ρ = sup n ( 1 − Q n + 1 ) . Since 0 ≤ Q n + 1 ≤ ( l ~ λ ˉ T ) n + 1 / ( n + 1 ) ! 0\le Q_{n+1}\le(\tilde{l}\bar\lambda T)^{n+1}/(n+1)! 0 ≤ Q n + 1 ≤ ( l ~ λ ˉ T ) n + 1 / ( n + 1 )! by claim 2, and the partial sums s n s_n s n of the series ∑ m ( l ~ λ ˉ T ) m / m ! = exp ( l ~ λ ˉ T ) \sum_m(\tilde{l}\bar\lambda T)^{m}/m!=\exp(\tilde{l}\bar\lambda T) ∑ m ( l ~ λ ˉ T ) m / m ! = exp ( l ~ λ ˉ T ) converge (The Real Exponential Function ), so that their differences s n + 1 − s n = ( l ~ λ ˉ T ) n + 1 / ( n + 1 ) ! s_{n+1}-s_n=(\tilde{l}\bar\lambda T)^{n+1}/(n+1)! s n + 1 − s n = ( l ~ λ ˉ T ) n + 1 / ( n + 1 )! tend to 0 0 0 by claim 1 of Arithmetic of Limits of Real Sequences , we get Q n + 1 → 0 Q_{n+1}\to0 Q n + 1 → 0 and ∫ R ℓ λ d ρ = 1 \int_{\mathbf{R}}\ell_\lambda\,d\rho=1 ∫ R ℓ λ d ρ = 1 . ■ \blacksquare ■