An Itō formula in S via Itō s Regularization

Size: px
Start display at page:

Download "An Itō formula in S via Itō s Regularization"

Transcription

1 isibang/ms/214/3 January 31st, statmath/eprints An Itō formula in S via Itō s Regularization Suprio Bhar Indian Statistical Institute, Bangalore Centre 8th Mile Mysore Road, Bangalore, 5659 India

2

3 AN ITŌ FORMULA IN S VIA ITŌ S REGULARIZATION SUPRIO BHAR Abstract. Following the work of Itaru Mitoma in [8], we give an alternative construction of Stochastic Integrals of Hermite Sobolev space valued predictable processes {G t } with respect to real semimartingales {X t }. We consider Gs, ϕ dxs for ϕ in the Schwartz space as random linear functionals. Our main technique involves obtaining regularized versions of these random linear functionals via Itō s Regularization Theorem. These results allow us to prove an Itō formula in S. 1. Introduction In various situations we come across random evaluation functionals. Given a real valued random variable X and a space E of real valued test functions, f f(x) for f E is an example of such a functional. Formally writing f(x) = δ X, f (where δ X takes values in the dual of E, some suitable space of distributions) gives a much clearer picture of this setup. More generally, we have random linear functionals which may take values in some space of distributions. Note that δ X actually takes values in some Hilbert space (see [11, Lemma 2.1]). Likewise if we can identify some regularity properties of a random linear functional (such as it takes values in a Hilbert space), analysis of such a functional becomes easier. First we use the setup of random linear functionals on Schwartz space to define Stochastic Integrals of Hermite-Sobolev space valued processes with respect to real semimartingales. Our main tool is Itō s Regularization Theorem which relates random linear functionals on the Schwartz space to processes taking values in some Hermite-Sobolev space. Using this identification, we then go on to prove an Itō formula. In section 2, we recall the countably Hilbertian topology defined on S(R d ) which gives rise to the Hermite-Sobolev spaces S p (R d ). We also recall a version of Itō s Regularization Theorem, which we use thoughout this paper. In section 3, we give a construction of the stochastic integral of appropriate predictable processes (taking values in the Hermite-Sobolev spaces) with respect to real square integrable martingales, finite variation processes and real semimartingales. We start with a Hermite-Sobolev space valued predictable processes {G t } and a real square integrable martingale {M t } and consider the random linear functional defined on the Schwartz space, viz. ϕ G s, ϕ [ dm s. In [8, section 3, Application ] t IV] the author remarked upon the condition E G s, ξ 2 d M s <, ξ E 21 Mathematics Subject Classification. Primary: 6H5; Secondary: 6G2. Key words and phrases. Hermite-Sobolev spaces, Tempered distributions, S valued processes, Random Linear functionals, Itō formula, Local times, Stochastic Integral, Tensor Quadratic Variation. 1

4 2 SUPRIO BHAR (E is some Nuclear Fréchet space) to obtain a E valued [ version of the stochastic integral. We consider a stronger assumption that E ] t G s 2 p d M s <, where G s takes values in some Hermite Sobolev space S p. This allows us to construct the version in S p, which we also connect to the theory of Hilbert space valued stochastic integrals as developed in [7]. We also compute the Tensor Quadratic Variation of the stochastic integral. Next we consider a norm bounded S p valued predictable process {G t } and produce a regularized version of the random linear functional ϕ G s, ϕ da s, where {A t } is some real valued process of locally finite variation. We show that the regularized version is a S p valued process of locally finite variation. We then extend this construction by considering a real valued semimartingale {X t } as the integrator. We use the decomposition of X into a martingale part M and a finite variation part A, to decompose the random linear functional ϕ G s, ϕ dx s as ϕ G s, ϕ dm s + G s, ϕ da s. We show that the regularized version of this random linear functional does not depend upon the decomposition M + A. We have followed the procedure for general stochastic integration as set out here [5, Chapter 23]. The results are extendable to multi dimensions. In section 4, we prove an Itō formula (see Theorem 4.2) which generalises [11, Theorem 2.3]. This result has been used in [12] to show existence of solution of some stochastic differential equations in S, the space of tempered distributions. A version of this Itō formula was also proved in [19, Theorem III.1] with equality in S. In [6, Theorem 3], the author has proved this formula for twice continuously (Fréchet) differentiable function while dealing with a single Hilbert space. Note that derivatives of tempered distributions may not not be in the same Hermite Sobolev space as the original one. Using regularization, the result [6, Theorem 3] was also proved in [8, Theorem 8] in the case of a E valued continuous martingale. These results are also connected to [11, Proposition 1.3], where the random linear functional nature of the stochastic integral has been pointed out. We also mention some connections with the Local Time process for a continuous real valued semimartingale (also see [11, Lemma 2.1]). The author would like to inform the interested reader that the general case of regularizing S valued semimartingales has been considered in [1]. 2. Preliminaries 2.1. Topologies on S and S. Let Z d + := {n = (n 1,, n d ) : n i integers}. If n = (n 1,, n d ), we define n := n n d. Let {h n : n Z d +} be the orthonormal basis (ONB) for L 2 (R d ) given by the Hermite functions (see [4, Chapter 1.3], [17, Chapter 1]). Let, represent the L 2 (R d ) inner product. For any fixed p R, consider the following formal sums { f, g p := k= n =k (2.1) (2k + d)2p f, h n g, h n, f 2 p := k= n =k (2k + d)2p f, h n 2 Let S(R d ) L 2 (R d ) be the space of smooth rapidly decreasing R-valued functions on R d with the topology given by L. Schwartz (see [18]). Then (S(R d ), p ) are pre-hilbert spaces and completing them one obtains the separable Hilbert spaces

5 AN ITŌ FORMULA IN S VIA ITŌ S REGULARIZATION 3 (S p (R d ), p ) known as the Hermite-Sobolev spaces (see [4, Chapter 1.3]). Let S (R d ) denote the space of tempered distributions. In [4], it was shown that (S p (R d ), p ) are dual to (S p (R d ), p ). Furthermore, the following are also known: L 2 (R d ) = (S (R d ), ), for p < q, (S q (R d ), q ) (S p (R d ), p ), S(R d ) = p R (S p(r d ), p ), S (R d ) = p R (S p(r d ), p ) The following notations will be used thoughout this paper: S, S, S p will stand for S(R), S (R), S p (R) respectively. Given ψ S(R d ) (or S p (R d )) and ϕ S (R d ) (or S p (R d )), the action of ϕ on ψ will be denoted by ϕ, ψ Regularization of random linear functionals. Definition 2.1. Let E be a countably Hilbertian Nuclear space. A stochastic process {X(x), x E} is a random linear functional if for each x, y E and a, b R we have X(ax + by) = ax(x) + bx(y) a.s. We shall take S(R d ) as our Nuclear space. The main tool for our analysis is Itō s Regularization Theorem for S(R d ). The result is as follows: Theorem 2.2 ([4, Theorem 2.3.2]). Let τ denote the topology on S(R d ). If Y is a τ-continuous random linear functional on S(R d ), then there exists a τ-regular version of Y say Ỹ, i.e. (1) Y (ϕ) = Ỹ, ϕ a.s. for all ϕ S(R d ). (2) There exists a countably Hilbertian topology θ on S(R d ) weaker than τ such that Ỹ is almost surely (S(Rd ), θ) valued. Furthermore such a τ-regular version is unique in the sense that if Ỹ and Y are two τ-regular version of Y, then Ỹ = Y a.s. For our purpose we need a specialized version of the above Theorem. Theorem 2.3 ([2, Theorem 4.1]). Let X be a random linear functional on S(R d ) which is continuous in probability in p for some p R. If p HS q, for some q R, then X has a version which is in S q (R d ) a.s. Remark 2.4. Note that in dimension 1, h q n := (2n + 1) q h n is an orthonormal basis for S q. Then, h q n 2 p = (2n + 1) 2(p q). n= n= Now for p HS q we need the above sum to be finite which happens if q > p For dimension d, the appropriate condition is q > p + d 2. Here is an outline of the proof. We have k= n =k hq n 2 p k= (2k +d)2(p q) #{n Z d + : n = k}. By [2, Chapter II, section 5], #{n Z d + : n = k} = ( ) k+d 1 d 1. We can find a constant C > such that ( ) k+d 1 d 1 C.(2k + d) d 1. Hence for q > p + d 2, the series is finite.

6 4 SUPRIO BHAR Remark 2.5. Continuity in probability can be metrized by the well-known Ky Fan metric. (2.2) d(y 1, Y 2 ) := E{ Y 1 Y 2 1} where Y 1, Y 2 are two real-valued random variables. Now, given a random linear functional X on S, to check the continuity property, we only need to check it at the point. Continuity in probability in some norm p means the following: given any ϵ >, δ > such that Since X() = we need to check x p < δ = d(x(x), X()) < ϵ, (2.3) x p < δ = E ( X(x) 1) < ϵ. Observe that (2.4) E ( X(x) 1) E ( X(x) ) ( E (X(x)) 2) 1 2 Mostly we shall estimate this second moment to establish the required continuity. 3. Regularized Version of the Stochastic Integral Unless stated otherwise the we shall use the following notations thoughout the section. To avoid cumbersome notations we mostly present the version of the results in dimension d = 1. We also follow the procedure for general stochastic integration as set out here [5, Chapter 23]. Fix p, q R with p > and q > p Note that this implies p HS q. Let T be a positive real number or. Our time interval will be [, T ]. For T =, [, T ] will mean [, ). Let (Ω, F, {F t } t [,T ], P ) be a probability space, with the filtration satisfying the usual conditions. For any real valued martingale M or a process of finite variation A or a semimartingale X, we shall assume M, A, X. Furthermore, any such real valued process (martingales, process of finite variation or semimartingales) will be assumed to have rcll (right continuous with left limits) paths. Unless stated otherwise stopping times or processes will be assumed to be adapted to the filtration {F t } Stochastic Integral with respect to a real square integrable Martingale. Let {M t } t [,T ] be a real-valued square integrable martingale adapted to the given filtration. Let {G t } t [,T ] be a S p valued predictable process such that (3.1) E T G s 2 p d M s <, Here M is the predictable increasing process such that M 2 M is a F t martingale. Note that for any ϕ S p, the real valued process G s, ϕ is a predictable process. Furthermore, we have for any t [, T ] (3.2) E G s, ϕ 2 d M s ϕ 2 p E G s 2 p d M s < Then for any t [, T ] we may consider the linear map I t,g on S p defined by (3.3) I t,g (ϕ) := G s, ϕ dm s.

7 AN ITŌ FORMULA IN S VIA ITŌ S REGULARIZATION 5 We can restrict this map to S and obtain a random linear functional on S. Theorem 3.1. For each t [, T ], the random linear functional I t,g has a regularized version Ĩt,G in S p. Proof. For any t [, T ], we have E(I t,g (ϕ)) 2 = E ϕ 2 p E ϕ 2 p E G s, ϕ 2 d M s, (by Itō s isometry) T G s 2 pd M s G s 2 pd M s Now using (2.4) we have ϵ ϕ p < 1 + E = E( I T t,g(ϕ) 1) < ϵ. G s 2 pd M s Hence I t,g is continuous in probability in p at ϕ =, which by linearity of I t,g is true globally. Hence by Theorem 2.3, we get a S q valued regularized version, say Ĩt,G, such that Ĩt,G, ϕ = I t,g (ϕ), a.s., ϕ S. Now another application of Itō s isometry gives ( ) 2 > E G s 2 p d M s = (2n + 1) 2p E G s, h n dm s = E n= 2 (2n + 1) 2p Ĩ t,g, h n = E Ĩ t,g 2 p. n= Hence Ĩt,G is S p valued a.s. The next result is an application of the well-known Pettis Theorem (see [15, Theorem 1.1]) Lemma 3.2. Let t [, T ] and p R. Let Y : Ω S p be a function. Then Y is measurable with respect to F t and B(S p ) (the Borel σ algebra on S p ) iff for all ϕ S, Y, ϕ is measurable with respect to F t and B(R) (the Borel σ algebra on R). Proof. For ϕ S, the map ψ ψ, ϕ is continuous on S p. If Y is measurable with respect to F t and B(S p ), then Y, ϕ is measurable as in the statement. For the converse, note that S p is a separable Hilbert space and its dual is S p. Furthermore, S is dense in S p in the norm p. If for all ϕ S, Y, ϕ is measurable with respect to F t and B(R), then by above density, we have Y, ϕ is measurable with respect to F t and B(R) for all ϕ S p. Applying Pettis Theorem we get the converse implication. We list properties of Ĩt,G in the next result. Theorem 3.3. Ĩt,G satisfies the following conditions:

8 6 SUPRIO BHAR (a) E Ĩt,G 2 p <. (b) Ĩt,G is measurable with respect to F t and the Borel σ-algebra of S p. (c) For any s, t [, T ] with s t and A F s, ] ] E [Ĩt,G 1 A = E [Ĩs,G 1 A. Hence Ĩt,G is a S p valued square integrable martingale. Proof. Part (a) is included in the proof of the previous Theorem. Note that Ĩt,G, ϕ = I t,g (ϕ), a.s., ϕ S. But Ĩt,G is S p valued and I t,g (ϕ) makes sense for all ϕ S p. Since S p is dual to S p and S is dense in S p we have Ĩt,G, ϕ = I t,g (ϕ), a.s., ϕ S p. Now I t,g (ϕ) is F t /B(R) measurable for all ϕ S and hence for all ϕ S p. Therefore Ĩt,G, ϕ is F t /B(R) measurable for all ϕ S p. Applying Lemma 3.2 we get part (b). For part (c), it is enough to prove (see discussion in [16, pp ]) ] ] E [Ĩs,G 1 A, ϕ = E [Ĩt,G 1 A, ϕ, ϕ S p. But, {I t,g (ϕ)} t is a martingale for any ϕ S. Hence, we have the following implications: E [I s,g (ϕ)1 A ] = E [I t,g (ϕ)1 A ], ϕ S. ] ] = E [ Ĩs,G, ϕ 1 A = E [ Ĩt,G, ϕ 1 A, ϕ S. = E [ Ĩs,G 1 A, ϕ ] = E [ Ĩt,G 1 A, ϕ ], ϕ S. ] ] = E [Ĩs,G 1 A, ϕ = E [Ĩt,G 1 A, ϕ, ϕ S. ] ] = E [Ĩs,G 1 A, ϕ = E [Ĩt,G 1 A, ϕ, ϕ S p. This proves part (c). Proposition 3.4. {Ĩt,G} has a version with rcll (right continuous with left limits) paths in q for any q > p Proof. Note that E(I t,g (ϕ)) 2 ϕ 2 p E T G s 2 pd M s. From [9, Theorem 1] we get the required version. Remark 3.5. Observe that for predictable rectangles G t = 1 (s,u] F (t, ω)ϕ, where s, u [, T ] with < s < u and ϕ S p, we have Ĩt,G = 1 F (M t u M t s )ϕ, which agrees with the usual stochastic integral defined on S p (as a Hilbert space, see [7]). Therefore the construction above actually gives the stochastic integral G s dm s for predictable G satisfying the integrability condition (3.1). Remark 3.6. We list some general observations.

9 AN ITŌ FORMULA IN S VIA ITŌ S REGULARIZATION 7 (1) If the martingale is continuous, then the argument works for integrands which are progressively measurable. In this case, we can replace M (in the integrability conditions) by [M], the quadratic variation process. (2) If M t = (Mt 1,, Mt n ) is a F t -adapted R n -valued square integrable martingale and {G t } is a S p (R d ) valued predictable process then under appropriate integrability conditions we may define the random linear functionals on S(R d ) I t,g,i (ϕ) := G s, ϕ dm i s, ϕ S(R d ); i = 1,, n. We can obtain regularized versions of the integrals. (3) we can do this procedure with locally square integrable martingales. The linear functionals need to be appropriately defined. The predictable process {G t } can also be taken to satisfy integrability condition (3.1) locally. Let B be a real Banach space and H be a real Hilbert space. Let B denote the norm in B. Let H and, H denote the norm and the inner product in H respectively. (see [7, Chapter 4, Section 21]) Definition 3.7. Let u B B. We define a norm on B B by { n } n u 1 := inf x i B y i B : u = x i y i, with x i, y i B. The projective tensor product B 1 B is the completion of B B with respect to 1. Given elements u, v of H H with u = n x i y i, v = m j=1 x j y j where x i, x j, y i, y j H, we define an inner product n m u, v 2 := xi, x j yi, y j j=1 The Hilbert-Schmidt tensor product denoted by H 2 H is the completion of H H with respect to this inner product. We also recall the definition of the tensor quadratic variation of a Hilbert space valued martingale (see [7, Theorem 21.6 and 22.8]). Definition 3.8 (Tensor Quadratic Variation). Given a H valued square integrable martingale, its tensor quadratic variation denoted by {[[M]] t } is the process [[M]] t = lim prob. P, P={=t <t 1 < <t n =t} H H. n 1 (M ti+1 M ti ) 2, where the limit (in probability) is of H 2 H valued random variables and P is the maximum length of the sub-intervals in P. We denote by { M t } the dual predictable projection of the tensor quadratic variation. Proposition 3.9. The tensor quadratic variation of Ĩt,G is given by ]] [[Ĩ,G = t i= G s G s d [M] s, a.s.

10 8 SUPRIO BHAR and its dual predictable projection is Ĩ,G t = G s G s d M s, a.s. Proof. Given any partition P = { = t < t 1 < < t n = t} and ϕ, ψ S, we have n 1 ( Iti+1,G(ϕ) I ) ( ti,g(ϕ) I ti+1,g(ψ) I ) ti,g(ψ) i= n 1 = (Ĩt i+1,g Ĩt i,g) 2, ϕ ψ, a.s. = i= n 1 i= ]] [[Ĩ,G (Ĩt i+1,g Ĩt i,g) 2, ϕ ψ, a.s. Since both and [I,G(ϕ), I,G (ψ)] t exist as appropriate limits (in probability) over shrinking partitions, we have t ]] (3.4) [I,G (ϕ), I,G (ψ)] t = [[Ĩ,G, ϕ ψ, a.s. Now [I,G (ϕ), I,G (ψ)] t = G s, ϕ G s, ψ d [M] s = t G 2 s, ϕ ψ d [M] s, a.s. Note that E G s 2 S p d [M] s = E G s 2 S p d M s is finite and G s 2 S p = G s G s 2. Hence we can write (using the theory of Bochner-Stieltjes integration) [I,G (ϕ), I,G (ψ)] t = G 2 s d [M] s, ϕ ψ, a.s. From (3.4) we now can identify process. ]] [[Ĩ,G t as the indicated S p 2 S p valued Since G s G s 1 = G s 2 S p = G s G s 2, then from (3.1) we get T E G s G s 1 d M s <. Now G s G s is a S p 1 S p valued predictable process. Therefore G s G s d M s is a S p 1 S p valued predictable process (using the theory of Bochner-Stieltjes integration). Note that H 1 H H 2 H. Claim: {Ĩt,G Ĩt,G } t G s G s d M s is a S p 1 S p valued martingale. ]] } Assuming the claim, since {Ĩt,G Ĩt,G [[Ĩ,G is a S p 1 S p valued martingale (see [7, Theorem 22.8(2 )]), then taking the difference we get { [[Ĩ,G]] } G s G s d M s t t

11 AN ITŌ FORMULA IN S VIA ITŌ S REGULARIZATION 9 is a S p 1 S p valued martingale. Since { G s G s d M s } is a predictable process, this proves that (also see [7, Proposition 22.7, equation (22.7.5)]) Ĩ,G = t G s G s d M s, a.s. Now we are going to prove the claim. For ϕ S we have by the Itō isometry ( ) 2 E Ĩt,G, ϕ = E(I t,g (ϕ)) 2 = E G s, ϕ 2 d M s or ) E ( Ĩt,G Ĩt,G, ϕ ϕ = E G s G s, ϕ ϕ d M s Given ϕ, ψ S, we can apply the previous equation to ϕ+ψ which on simplification gives ) (3.5) E ( Ĩt,G Ĩt,G, ϕ ψ = E G s G s, ϕ ψ d M s Again from Theorem 3.3(a) we get E Ĩt,G 2 S p < which implies E Ĩt,G Ĩt,G 1 <. Rewriting equation (3.5) we get ) E (Ĩt,G Ĩt,G, ϕ ψ = E G s G s d M s, ϕ ψ which implies (3.6) E (Ĩt,G Ĩt,G ) == E S p S p G s G s d M s Now fix u, t [, T ] with u t. Let F F u. Since I t,g (ϕ) = G s, ϕ dm s is a martingale, we have E [ [ (I t,g (ϕ)) 2 ] t ] 1 F = E 1 F G s, ϕ 2 d M s. Following the same derivation as in (3.6) we get ) ] ] (3.7) E [(Ĩt,G Ĩt,G 1 F == E [1 F G s G s d M S s p S p This proves the claim Stochastic Integral with respect to a real finite variation process. Let {A t } t [,T ] be a real valued F t -adapted process of finite variation with right continuous paths such that its total variation on the time interval [, T ] is bounded by a constant, i.e. C > such that (3.8) V [,T ] (A ) C, a.s. Note that if T =, then by (3.8) we mean that for any t >, V [,t] (A ) C a.s. Let {G t } t [,T ] be a S p valued norm bounded (i.e. there exists a constant R > such that G s p R a.s. for all t) predictable process. In particular we have for any t [, T ] and ϕ S p, (3.9) G s, ϕ G s p ϕ p R. ϕ p

12 1 SUPRIO BHAR and (3.1) G s p da s R.V [,t] (A ) R.C <. This also implies G s, ϕ da s R.C. ϕ p. Now consider the linear map J t,g on S p defined by (3.11) J t,g (ϕ) := G s, ϕ da s. We can restrict this map to S and obtain a random linear functional on S. Theorem 3.1. For each t [, T ], the random linear functional J t,g has a regularized version J t,g in S p. Proof. Note that Then, E (J t,g (ϕ) R.C ϕ p. So J t,g (ϕ) ϕ p G s p da s RC ϕ p. ϕ p < ϵ R.C = E( J t,g(ϕ) 1) < ϵ. Hence J t,g is continuous in probability in p at ϕ =, which by linearity of J t,g is true globally. Hence by Theorem 2.3, we get a S q valued regularized version, say J t,g, such that Jt,G, ϕ = J t,g (ϕ), a.s., ϕ S. Note that if for some ω Ω, V [,t] A (ω) =, then (J t,g (ϕ))(ω) =, which does not contribute to E J t,g (ϕ). For further calculations we shall ignore these ω. Now, E J t,g 2 p = E = = 2 (2n + 1) 2p Jt,G, h n n= (2n + 1) 2p E (J t,g (h n )) 2 n= ( (2n + 1) 2p E n= ) 2 G s, h n da s ( (2n + 1) 2p da s E G s, h n V [,t] (A ) V [,t] (A ) n= using Jensen s inequality, ( (2n + 1) 2p E n= ) 2 ) G s, h n 2 (V [,t] (A )) 2 da s V [,t] (A )

13 AN ITŌ FORMULA IN S VIA ITŌ S REGULARIZATION 11 using (3.8), C.E ( = C. E <. ( ) (2n + 1) 2p G s, h n 2 da s n= ) G s 2 p da s C.R 2 ( E V [,t] (A ) ) Hence J t,g is S p valued a.s. Proposition Jt,G has the following properties: (a) J t,g is measurable with respect to F t and the Borel σ-algebra of S p. (b) J t,g is a S p valued finite variation process. Proof. Proof of part (a) is similar to Theorem (3.3) part (b). To prove part (b), it is enough to show the variation exists for time intervals of the form [, t] where t [, T ]. Fix such a t. Let P := { = t < t 1 < t n = t} be a partition of [, t]. Fix a ω Ω. Take any arbitrary ϵ >. Then there exist ϕ, ϕ 1,, ϕ n 1 S (all of which depends on ω) such that ϕ i p 1 for i =, 1,, n 1 and J ti+1,g J ti,g p (ω) Then, n 1 J ti+1,g J n 1 ti,g p (ω) i= Using (3.1), we have Jti+1,G J ti,g, ϕ i (ω) + i= Jti+1,G J n 1 ti,g, ϕ i (ω) + ϵ, i =, 1,, n 1. 2i+1 i= n 1 < ( J ti+1,g(ϕ i ) J ti,g(ϕ i ) ) (ω) + ϵ i= n 1 i= i+1 t i G s, ϕ i da s (ω) + ϵ ϵ 2 i+1 n 1 (i+1 ) ϕ i p G s p da s (ω) + ϵ i= t i n 1 (i+1 ) G s p da s (ω) + ϵ, ( ϕ i p 1) i= t i ( ) = G s p da s (ω) + ϵ n 1 J ti+1,g J ti,g p < RC + ϵ, a.s. i= But P is an arbitrary partition of [, t] and the constant RC does not depend on such a partition. Hence taking supremum over all such partitions we conclude that the variation is finite a.s.

14 12 SUPRIO BHAR Proposition Jt,G has rcll (right continuous with left limits) paths in q for any q > p Proof. Note that sup s [,t] J s,g (ϕ) ϕ p G s p da s RC ϕ p. The arguments are same as in [9, Theorem 1]. Remark Observe that for predictable rectangles G t = 1 (s,u] F (t, ω)ϕ, where s, u [, T ] with < s < u and ϕ S p, we have J t,g = 1 F (A t u A t s )ϕ, which agrees with the Bochner integral defined on S p (in the Stieltjes sense). Therefore the construction above actually gives the stochastic integral G s da s for normbounded and predictable G. The construction given above can be extended to any finite variation process. Proposition Let {A t } t [,T ] be a F t -adapted process of finite variation with right continuous paths. Let {G t } t [,T ] be a S p valued norm bounded [i.e. there exists a constant R > such that G s p R a.s. for all t] predictable process. Then there exist a sequence of F t stopping times τ n with τ n and a sequence of (n) S p valued regularized versions J t,g of J t,g (as defined in (3.11)) such that J (n) (a) t,g is measurable with respect to F t and the Borel σ-algebra of S p. (n) (b) J t,g is a S p valued finite variation process. (n) (c) For t [, τ n ) and for all ψ S, J t,g, ψ = J t,g (ψ) a.s.. Proof. Note that V [,t] (A ) is a non-decreasing process adapted to the filtration. Consider the stopping times τ n defined by τ n := inf{t : V [,t] (A ) n}. Note that there might be jumps of A (and hence of V [,t] (A )) at the time points t = τ n (ω). But these are prelocally bounded. In particular, the processes A τ n t := A t 1 (t<τn) + A τn 1 (τn t) are F t adapted and bounded. Furthermore we actually have V [,t] (A τ n ) n. Now consider random linear functionals J (n) t,g (ϕ) := G s, ϕ da τ n s. Note that J (n) t,g (ϕ) = J t,g(ϕ) a.s. for t < τ n. We can obtain regularized versions of J (n) t,g (ϕ) (as in Theorem 3.1 and Proposition 3.11), which gives the appropriate versions mentioned in the statement. Remark If the paths of the finite variation process are continuous, then so are the paths of the total variation. In this situation we may take t [, τ n ] in part (c) of Proposition 3.14 (since there are no jumps at t = τ n ). Remark We can relax the hypothesis using the following: (1) Given a process of locally finite variation, by stopping we get a process of finite variation and hence the procedure given above works for processes of locally finite variation. The predictable process {G t } can also be taken to be locally norm bounded.

15 AN ITŌ FORMULA IN S VIA ITŌ S REGULARIZATION 13 (2) We can take {G t } to be S p (R d ) valued. In which case the random linear functional J t,g (ϕ) needs to be defined for ϕ S(R d ). All the results mentioned above can be proved in this general setup. (3) If A t = (A 1 t,, A n t ) is a F t -adapted R n -valued process of finite variation with right continuous paths and {G t } a S p (R d ) valued predictable process then under appropriate integrability conditions we may define the random linear functionals on S(R d ) J t,g,i (ϕ) := G s, ϕ da i s, ϕ S(R d ); i = 1,, n. We can obtain regularized versions of the integrals Stochastic Integral with respect to a real semimartingale. A important step in defining stochastic integrals of real valued bounded predictable processes with respect to a real semimartingale is the following: Lemma Let {V t } t be a real valued bounded predictable processes. Let {M t } t be a square integrable martingale and {A t } t is a process of finite variation such that M t = A t, a.s. t. Then V s dm s = V s da s, a.s. t. Proof. This result is included in the proof of Theorem 23.4 in [5]. A classical result on local martingales goes like this: (see [5, Lemma 23.5] or [1, Theorem 25]) Theorem 3.18 (Decomposition of Local Martingales). Given any local martingale {M t } t, two local martingales {M t} t, {M t } t one of which has bounded jumps and the other is of locally integrable variation and M t = M t + M t, a.s. Thus any semimartingale has a decomposition (not necessarily unique) X t = M }{{} t + A }{{} t, a.s. locally square integrable locally finite variation Suitably stopping, we can assume them to be a square integrable martingale and a process of finite variation respectively. Again using further stopping, the process of finite variation can be made to satisfy (3.8) prelocally (see Proposition 3.14). Theorem Let {X t } be a real semimartingale such that it has a decomposition X t = M t + A t, a.s. t, where {M t } t is a square integrable martingale and {A t } t is a process of finite variation satisfying (3.8). Let {G t } t be a norm bounded predictable S p valued process (i.e. G t p is bounded). Then G s, ϕ dx s has a regularized version in S p. Furthermore, this version is independent of the decomposition X = M + A.

16 14 SUPRIO BHAR Proof. Under the hypothesis the integrability conditions (3.1) and (3.1) are satisfied on any finite time interval [, T ]. Then, for any t [, T ] and ϕ S, we have Hence (Ĩt,G + J ) t,g G s, ϕ dx s = G s, ϕ dm s + G s, ϕ da s }{{}}{{} I t,g (ϕ) J t,g (ϕ) = Ĩt,G, ϕ + Jt,G, ϕ, a.s. gives a regularized version of the integral G s, ϕ dx s. Apriori, it seems that this version will depend on the decomposition of X. Now consider the real bounded predictable process V s = G s, ϕ. Hence Lemma 3.17 gives, G s, ϕ dm s = G s, ϕ da s, a.s. if M = A with M a square integrable martingale and A a process of finite variation. This says the stochastic integral G s, ϕ dx s does not depend on the decomposition of X. We apply the same result here to conclude that the regularized version is indeed independent of the decomposition. Remark 3.2. As mentioned before in the remarks in previous subsections, we may also consider the case where {G t } is a S p (R d ) valued locally norm-bounded predictable process and X t = (X 1 t,, X n t ) is a R n valued semimartingale, to obtain regularized versions of stochastic integrals under appropriate integrability conditions. 4. Itō Formula Consider the derivative maps denoted by i : S(R d ) S(R d ) for i = 1,, d. We can extend these maps by duality to i : S (R d ) S (R d ) as follows: for ψ S (R d ), i ψ, ϕ := ψ, i ϕ, ϕ S(R d ). Let {e i : i = 1,, d} be the standard basis vectors in R d. Let n = (n 1,, n d ) Z d + be a multi-index. Then (see [3, Appendix A.5]) ni e i ni + e i i h n = h n ei h n+ei, 2 2 with the convention that for a multi-index n = (n 1,, n d ), if n i < for some i, then h n. Above recurrence implies that i : S p (R d ) S p 1 (Rd ) is a 2 bounded linear operator. First we recall some useful facts about translation operators on S (R d ). For x R d, let τ x denote the operators on S(R d ) defined by (τ x ϕ)(y) := ϕ(y x), y R d. This operators can be extended to τ x : S (R d ) S (R d ) by τ x ϕ, ψ := ϕ, τ x ψ, ψ S(R d ). Proposition 4.1. The translation operators defined have the following properties:

17 AN ITŌ FORMULA IN S VIA ITŌ S REGULARIZATION 15 (a) For x R d and any p R, τ x : S p (R d ) S p (R d ) is a bounded linear map. In particular, there exists a real polynomial P k of degree k = 2( p + 1) such that τ x ϕ p P k ( x ) ϕ p, ϕ S p (R d ). (b) For x R d and any i = 1,, d we have τ x i = i τ x. Proof. See [13, Theorem 2.1] for the proof of part (a). Fix an element ψ S(R d ). Then for y R d, (τ x i ψ)(y) = ( i ψ)(y + x) = i (ψ(y + x)) = ( i τ x ψ)(y) i.e. τ x i ψ = i τ x ψ. Now for ϕ S (R d ) and any ψ S(R d ), Hence we have part (b). i τ x ϕ, ψ = τ x ϕ, i ψ = ϕ, τ x i ψ = ϕ, i τ x ψ = i ϕ, τ x ψ = τ x i ϕ, ψ. Given a ϕ S (R d ), there exists a p > such that ϕ S p (R d ). Let {X t } be a (R d ) valued semimartingale. By part (a) of the Proposition 4.1, we have τ Xt ϕ is S p (R d ) valued for any t. By stopping, if required, we may assume that {X t } is prelocally bounded. But i : S p (R d ) S p 1 (Rd ) is a bounded linear operator. 2 Hence, by part (a) of Proposition 4.1 we get i τ Xs ϕ p 1 2 C. τ X s ϕ p C.P k ( X s ) ϕ p C, s [, T ] where C, C > are appropriate constants. Similarly, there exists a constant C > such that ij τ Xs ϕ p 1 C, s [, T ]. By stopping, if required, we may assume that the total variations of [X i, X j ] c t are also bounded (see Proposition 3.14). Hence as per the results obtained in the previous section, the random linear functionals d It 1 (ψ) := i τ Xs ϕ, ψ dxs i and I 2 t (ψ) := d i,j=1 2 ij τ Xs ϕ, ψ d[x i, X j ] c s both have regularized versions which live in S p 1 (Rd ) and S 2 p 1 (R d ) respectively and are given by d iτ Xs ϕ dxs i and d i,j=1 2 ij τ X s ϕ d[x i, X j ] c s respectively. We now prove an Itō formula which generalises [11, Theorem 2.3]. This result has been used in [12] to show existence of solution of some stochastic differential

18 16 SUPRIO BHAR equations in S, the space of tempered distributions. A version of this Itō formula was also proved in [19, Theorem III.1] with equality in S. In [6, Theorem 3], the author has proved this formula for twice continuously (Fréchet) differentiable function while dealing with a single Hilbert space. Note that derivatives of tempered distributions may not not be in the same Hermite Sobolev space as the original one. Using regularization, the result [6, Theorem 3] was also proved in [8, Theorem 8] in the case of a E valued continuous martingale. Theorem 4.2. Let p > and ϕ S p (R d ). Let X = (X 1,, X d ) be a R d valued semimartingale. Let Xs i denote the jump of Xs. i Then [ ] d τ Xs ϕ τ Xs ϕ + ( Xs i i τ Xs ϕ) s t is a S p 1 (R d ) valued process and we have the following equality in S p 1 (R d ), (4.1) d τ Xt ϕ = τ X ϕ i τ Xs ϕ dxs i + 1 d 2 ijτ 2 Xs ϕ d[x i, X j ] c s i,j=1 + [ ] d τ Xs ϕ τ Xs ϕ + ( Xs i i τ Xs ϕ), a.s. s t Proof. By [11, Proposition 1.4], there exists some positive integer n such that the map x τ x ϕ S n is a C 2 map. For any fixed ψ S(R d ) we have x τ x ϕ, ψ is a C 2 map and i τ x ϕ, ψ = i ϕ, ψ( + x) = ϕ, i ψ( + x) = ϕ, τ x i ψ = i τ x ϕ, ψ etc. We follow the proof of [5, Theorem 23.7] and define B t := {x R d : x sup s t X s }. Then there exist constants C, C > such that d τ Xs ϕ τ Xs ϕ + ( Xs i i τ Xs ϕ), ψ s t = τ X s ϕ, ψ τ Xs ϕ, ψ d + i τ Xs ϕ, ψ Xs i s t = τ X s ϕ, ψ τ Xs ϕ, ψ d i τxs ϕ, ψ Xs i s t C. d X s 2 sup ijτ 2 y ϕ, ψ s t i,j=1 y B t C. d X s 2 sup ijτ 2 y ϕ p 1 ψ p+1 s t i,j=1 y B t C. ( ) X s 2 sup τ y ϕ p ψ p+1. y B s t t

19 AN ITŌ FORMULA IN S VIA ITŌ S REGULARIZATION 17 Since for each s [, t], the random variable τ Xs ϕ τ Xs ϕ + d ( Xi s i τ Xs ϕ) is S p 1 valued, in view of the above estimate, [ ] d τ Xs ϕ τ Xs ϕ + ( Xs i i τ Xs ϕ) s t is a S p 1 valued random variable. To complete the proof we need to verify the following equality in S p 1, [ ] d d τ Xs ϕ τ Xs ϕ + ( Xs i i τ Xs ϕ) = τ Xt ϕ τ X ϕ + i τ Xs ϕ dxs i s t 1 2 d i,j=1 2 ijτ Xs ϕ d[x i, X j ] c s It is enough to show that the action of both the sides on ψ S agree. Now applying the Itō formula (see [5, Theorem 23.7]) to this C 2 map x τ x ϕ, ψ we have d τ Xt ϕ, ψ = τ X ϕ, ψ i τ Xs ϕ, ψ dxs i (4.2) } {{ } =It 1(ψ) + 1 d 2 2 ij τ Xs ϕ, ψ d[x i, X j ] c s i,j=1 }{{} + s t [ =I 2 t (ψ) τ Xs ϕ, ψ τ Xs ϕ, ψ + where X i s denotes the jump of X i s. This completes the proof. a.s. d i τ Xs ϕ, ψ ] Xs i, a.s., Given a real valued semimartingale {X t }, consider the local time process denoted by {L t (x)} t [, ),x R. Note that this process is jointly measurable in (x, t, ω) and for each x R, {L t (x)} t is a continuous adapted process. By [1, p. 216, Corollary 1], we have for any ϕ S, (4.3) L t (x)ϕ(x) dx = ϕ(x s )d [X, X] c s. As an application of Proposition 3.14 we have the following result. Proposition 4.3. The random tempered distribution δ X s d [X, X] c s is S p valued for any p > 1 4 and is given by the function x L t(x). Proof. Note that for any fixed x R, the distribution δ x is in S p for any p > 1 4 (see [14, Theorem 4.1 part (a)]). Also τ x δ = δ x. The map x τ x δ is continuous (see [14, proof of Proposition 3.1]). By stopping if required, we may assume {X t } is bounded. Hence {δ Xt } is a norm bounded predictable process. Hence the random linear functional ϕ ϕ(x s )d [X, X] c s = δxs, ϕ d [X, X] c s has a regularized version (see Proposition 3.14) given by

20 18 SUPRIO BHAR δ X s d [X, X] c s. By the occupation density formula (4.3) this random tempered distribution is given by the function x L t (x). Acknowledgement: The author would like to thank Professor B. Rajeev, Indian Statistical Institute, Bangalore for valuable suggestions during the work and pointing out the way to Theorem 4.2. References 1. S. Bhar, Regularization of S valued semimartingales, Preprint. 2. William Feller, An introduction to probability theory and its applications. Vol. I, Third edition, John Wiley & Sons Inc., New York, MR 2282 (37 #364) 3. Takeyuki Hida, Brownian motion, Applications of Mathematics, vol. 11, Springer-Verlag, New York, 198, Translated from the Japanese by the author and T. P. Speed. MR (81a:689) 4. Kiyosi Itō, Foundations of stochastic differential equations in infinite-dimensional spaces, CBMS-NSF Regional Conference Series in Applied Mathematics, vol. 47, Society for Industrial and Applied Mathematics (SIAM), Philadelphia, PA, MR (87a:668) 5. Olav Kallenberg, Foundations of modern probability, Probability and its Applications (New York), Springer-Verlag, New York, MR (99e:61) 6. Hiroshi Kunita, Stochastic integrals based on martingales taking values in Hilbert space, Nagoya Math. J. 38 (197), MR (41 #9345) 7. Michel Métivier, Semimartingales, de Gruyter Studies in Mathematics, vol. 2, Walter de Gruyter & Co., Berlin, 1982, A course on stochastic processes. MR (84i:62) 8. Itaru Mitoma, Martingales of random distributions, Mem. Fac. Sci. Kyushu Univ. Ser. A 35 (1981), no. 1, MR (82e:679) 9., On the norm continuity of S -valued gaussian processes, Nagoya Math. J. 82 (1981), MR (82h:67) 1. Philip E. Protter, Stochastic integration and differential equations, second ed., Applications of Mathematics (New York), vol. 21, Springer-Verlag, Berlin, 24, Stochastic Modelling and Applied Probability. MR (25k:68) 11. B. Rajeev, From Tanaka s formula to Ito s formula: distributions, tensor products and local times, 1755 (21), MR (22j:693) 12., Translation invariant diffusion in the space of tempered distributions, Indian J. Pure Appl. Math. 44 (213), no. 2, MR B. Rajeev and S. Thangavelu, Probabilistic representations of solutions to the heat equation, Proc. Indian Acad. Sci. Math. Sci. 113 (23), no. 3, MR (24g:61) 14., Probabilistic representations of solutions of the forward equations, Potential Anal. 28 (28), no. 2, MR (29e:6139) 15. R. E. Showalter, Monotone operators in Banach space and nonlinear partial differential equations, Mathematical Surveys and Monographs, vol. 49, American Mathematical Society, Providence, RI, MR (98c:4776) 16. Daniel W. Stroock, Probability theory, an analytic view, Cambridge University Press, Cambridge, MR (95f:63) 17. Sundaram Thangavelu, Lectures on Hermite and Laguerre expansions, Mathematical Notes, vol. 42, Princeton University Press, Princeton, NJ, 1993, With a preface by Robert S. Strichartz. MR (94i:421) 18. François Trèves, Topological vector spaces, distributions and kernels, Dover Publications Inc., Mineola, NY, 26, Unabridged republication of the 1967 original. MR (27k:462) 19. A. S. Üstünel, A generalization of Itô s formula, J. Funct. Anal. 47 (1982), no. 2, MR (83k:656) 2. J.B. Walsh, An introduction to stochastic partial differential equations, Lecture Notes in Mathematics, vol. 118, Springer, Suprio Bhar, Indian Statistical Institute Bangalore Centre. address: suprio@isibang.ac.in

Probabilistic representations of solutions to the heat equation

Probabilistic representations of solutions to the heat equation Proc. Indian Acad. Sci. (Math. Sci.) Vol. 113, No. 3, August 23, pp. 321 332. Printed in India Probabilistic representations of solutions to the heat equation B RAJEEV and S THANGAVELU Indian Statistical

More information

A note on the σ-algebra of cylinder sets and all that

A note on the σ-algebra of cylinder sets and all that A note on the σ-algebra of cylinder sets and all that José Luis Silva CCM, Univ. da Madeira, P-9000 Funchal Madeira BiBoS, Univ. of Bielefeld, Germany (luis@dragoeiro.uma.pt) September 1999 Abstract In

More information

GENERALIZED COVARIATION FOR BANACH SPACE VALUED PROCESSES, ITÔ FORMULA AND APPLICATIONS

GENERALIZED COVARIATION FOR BANACH SPACE VALUED PROCESSES, ITÔ FORMULA AND APPLICATIONS Di Girolami, C. and Russo, F. Osaka J. Math. 51 (214), 729 783 GENERALIZED COVARIATION FOR BANACH SPACE VALUED PROCESSES, ITÔ FORMULA AND APPLICATIONS CRISTINA DI GIROLAMI and FRANCESCO RUSSO (Received

More information

A Concise Course on Stochastic Partial Differential Equations

A Concise Course on Stochastic Partial Differential Equations A Concise Course on Stochastic Partial Differential Equations Michael Röckner Reference: C. Prevot, M. Röckner: Springer LN in Math. 1905, Berlin (2007) And see the references therein for the original

More information

Hilbert Spaces. Contents

Hilbert Spaces. Contents Hilbert Spaces Contents 1 Introducing Hilbert Spaces 1 1.1 Basic definitions........................... 1 1.2 Results about norms and inner products.............. 3 1.3 Banach and Hilbert spaces......................

More information

OPTIMAL SOLUTIONS TO STOCHASTIC DIFFERENTIAL INCLUSIONS

OPTIMAL SOLUTIONS TO STOCHASTIC DIFFERENTIAL INCLUSIONS APPLICATIONES MATHEMATICAE 29,4 (22), pp. 387 398 Mariusz Michta (Zielona Góra) OPTIMAL SOLUTIONS TO STOCHASTIC DIFFERENTIAL INCLUSIONS Abstract. A martingale problem approach is used first to analyze

More information

Independence of some multiple Poisson stochastic integrals with variable-sign kernels

Independence of some multiple Poisson stochastic integrals with variable-sign kernels Independence of some multiple Poisson stochastic integrals with variable-sign kernels Nicolas Privault Division of Mathematical Sciences School of Physical and Mathematical Sciences Nanyang Technological

More information

A NOTE ON STOCHASTIC INTEGRALS AS L 2 -CURVES

A NOTE ON STOCHASTIC INTEGRALS AS L 2 -CURVES A NOTE ON STOCHASTIC INTEGRALS AS L 2 -CURVES STEFAN TAPPE Abstract. In a work of van Gaans (25a) stochastic integrals are regarded as L 2 -curves. In Filipović and Tappe (28) we have shown the connection

More information

1. Stochastic Processes and filtrations

1. Stochastic Processes and filtrations 1. Stochastic Processes and 1. Stoch. pr., A stochastic process (X t ) t T is a collection of random variables on (Ω, F) with values in a measurable space (S, S), i.e., for all t, In our case X t : Ω S

More information

Banach Journal of Mathematical Analysis ISSN: (electronic)

Banach Journal of Mathematical Analysis ISSN: (electronic) Banach J. Math. Anal. 6 (2012), no. 1, 139 146 Banach Journal of Mathematical Analysis ISSN: 1735-8787 (electronic) www.emis.de/journals/bjma/ AN EXTENSION OF KY FAN S DOMINANCE THEOREM RAHIM ALIZADEH

More information

Citation Osaka Journal of Mathematics. 41(4)

Citation Osaka Journal of Mathematics. 41(4) TitleA non quasi-invariance of the Brown Authors Sadasue, Gaku Citation Osaka Journal of Mathematics. 414 Issue 4-1 Date Text Version publisher URL http://hdl.handle.net/1194/1174 DOI Rights Osaka University

More information

Brownian motion. Samy Tindel. Purdue University. Probability Theory 2 - MA 539

Brownian motion. Samy Tindel. Purdue University. Probability Theory 2 - MA 539 Brownian motion Samy Tindel Purdue University Probability Theory 2 - MA 539 Mostly taken from Brownian Motion and Stochastic Calculus by I. Karatzas and S. Shreve Samy T. Brownian motion Probability Theory

More information

Stochastic Processes II/ Wahrscheinlichkeitstheorie III. Lecture Notes

Stochastic Processes II/ Wahrscheinlichkeitstheorie III. Lecture Notes BMS Basic Course Stochastic Processes II/ Wahrscheinlichkeitstheorie III Michael Scheutzow Lecture Notes Technische Universität Berlin Sommersemester 218 preliminary version October 12th 218 Contents

More information

for all f satisfying E[ f(x) ] <.

for all f satisfying E[ f(x) ] <. . Let (Ω, F, P ) be a probability space and D be a sub-σ-algebra of F. An (H, H)-valued random variable X is independent of D if and only if P ({X Γ} D) = P {X Γ}P (D) for all Γ H and D D. Prove that if

More information

EULER MARUYAMA APPROXIMATION FOR SDES WITH JUMPS AND NON-LIPSCHITZ COEFFICIENTS

EULER MARUYAMA APPROXIMATION FOR SDES WITH JUMPS AND NON-LIPSCHITZ COEFFICIENTS Qiao, H. Osaka J. Math. 51 (14), 47 66 EULER MARUYAMA APPROXIMATION FOR SDES WITH JUMPS AND NON-LIPSCHITZ COEFFICIENTS HUIJIE QIAO (Received May 6, 11, revised May 1, 1) Abstract In this paper we show

More information

SUMMARY OF RESULTS ON PATH SPACES AND CONVERGENCE IN DISTRIBUTION FOR STOCHASTIC PROCESSES

SUMMARY OF RESULTS ON PATH SPACES AND CONVERGENCE IN DISTRIBUTION FOR STOCHASTIC PROCESSES SUMMARY OF RESULTS ON PATH SPACES AND CONVERGENCE IN DISTRIBUTION FOR STOCHASTIC PROCESSES RUTH J. WILLIAMS October 2, 2017 Department of Mathematics, University of California, San Diego, 9500 Gilman Drive,

More information

Finite-dimensional spaces. C n is the space of n-tuples x = (x 1,..., x n ) of complex numbers. It is a Hilbert space with the inner product

Finite-dimensional spaces. C n is the space of n-tuples x = (x 1,..., x n ) of complex numbers. It is a Hilbert space with the inner product Chapter 4 Hilbert Spaces 4.1 Inner Product Spaces Inner Product Space. A complex vector space E is called an inner product space (or a pre-hilbert space, or a unitary space) if there is a mapping (, )

More information

On Kusuoka Representation of Law Invariant Risk Measures

On Kusuoka Representation of Law Invariant Risk Measures MATHEMATICS OF OPERATIONS RESEARCH Vol. 38, No. 1, February 213, pp. 142 152 ISSN 364-765X (print) ISSN 1526-5471 (online) http://dx.doi.org/1.1287/moor.112.563 213 INFORMS On Kusuoka Representation of

More information

Brownian Motion. 1 Definition Brownian Motion Wiener measure... 3

Brownian Motion. 1 Definition Brownian Motion Wiener measure... 3 Brownian Motion Contents 1 Definition 2 1.1 Brownian Motion................................. 2 1.2 Wiener measure.................................. 3 2 Construction 4 2.1 Gaussian process.................................

More information

Generalized Gaussian Bridges of Prediction-Invertible Processes

Generalized Gaussian Bridges of Prediction-Invertible Processes Generalized Gaussian Bridges of Prediction-Invertible Processes Tommi Sottinen 1 and Adil Yazigi University of Vaasa, Finland Modern Stochastics: Theory and Applications III September 1, 212, Kyiv, Ukraine

More information

(B(t i+1 ) B(t i )) 2

(B(t i+1 ) B(t i )) 2 ltcc5.tex Week 5 29 October 213 Ch. V. ITÔ (STOCHASTIC) CALCULUS. WEAK CONVERGENCE. 1. Quadratic Variation. A partition π n of [, t] is a finite set of points t ni such that = t n < t n1

More information

Lecture 21 Representations of Martingales

Lecture 21 Representations of Martingales Lecture 21: Representations of Martingales 1 of 11 Course: Theory of Probability II Term: Spring 215 Instructor: Gordan Zitkovic Lecture 21 Representations of Martingales Right-continuous inverses Let

More information

Lecture 12. F o s, (1.1) F t := s>t

Lecture 12. F o s, (1.1) F t := s>t Lecture 12 1 Brownian motion: the Markov property Let C := C(0, ), R) be the space of continuous functions mapping from 0, ) to R, in which a Brownian motion (B t ) t 0 almost surely takes its value. Let

More information

Stochastic integration. P.J.C. Spreij

Stochastic integration. P.J.C. Spreij Stochastic integration P.J.C. Spreij this version: April 22, 29 Contents 1 Stochastic processes 1 1.1 General theory............................... 1 1.2 Stopping times...............................

More information

Stochastic Partial Differential Equations with Levy Noise

Stochastic Partial Differential Equations with Levy Noise Stochastic Partial Differential Equations with Levy Noise An Evolution Equation Approach S..PESZAT and J. ZABCZYK Institute of Mathematics, Polish Academy of Sciences' CAMBRIDGE UNIVERSITY PRESS Contents

More information

Regularity of the density for the stochastic heat equation

Regularity of the density for the stochastic heat equation Regularity of the density for the stochastic heat equation Carl Mueller 1 Department of Mathematics University of Rochester Rochester, NY 15627 USA email: cmlr@math.rochester.edu David Nualart 2 Department

More information

Solving Stochastic Partial Differential Equations as Stochastic Differential Equations in Infinite Dimensions - a Review

Solving Stochastic Partial Differential Equations as Stochastic Differential Equations in Infinite Dimensions - a Review Solving Stochastic Partial Differential Equations as Stochastic Differential Equations in Infinite Dimensions - a Review L. Gawarecki Kettering University NSF/CBMS Conference Analysis of Stochastic Partial

More information

ERRATA: Probabilistic Techniques in Analysis

ERRATA: Probabilistic Techniques in Analysis ERRATA: Probabilistic Techniques in Analysis ERRATA 1 Updated April 25, 26 Page 3, line 13. A 1,..., A n are independent if P(A i1 A ij ) = P(A 1 ) P(A ij ) for every subset {i 1,..., i j } of {1,...,

More information

CONTINUITY OF CP-SEMIGROUPS IN THE POINT-STRONG OPERATOR TOPOLOGY

CONTINUITY OF CP-SEMIGROUPS IN THE POINT-STRONG OPERATOR TOPOLOGY J. OPERATOR THEORY 64:1(21), 149 154 Copyright by THETA, 21 CONTINUITY OF CP-SEMIGROUPS IN THE POINT-STRONG OPERATOR TOPOLOGY DANIEL MARKIEWICZ and ORR MOSHE SHALIT Communicated by William Arveson ABSTRACT.

More information

Applications of Ito s Formula

Applications of Ito s Formula CHAPTER 4 Applications of Ito s Formula In this chapter, we discuss several basic theorems in stochastic analysis. Their proofs are good examples of applications of Itô s formula. 1. Lévy s martingale

More information

LOCALLY INTEGRABLE PROCESSES WITH RESPECT TO LOCALLY ADDITIVE SUMMABLE PROCESSES

LOCALLY INTEGRABLE PROCESSES WITH RESPECT TO LOCALLY ADDITIVE SUMMABLE PROCESSES LOCALLY INTEGRABLE PROCESSES WITH RESPECT TO LOCALLY ADDITIVE SUMMABLE PROCESSES OANA MOCIOALCA Abstract. In [MD] we defined and studied a class of summable processes, called additive summable processes,

More information

THE SKOROKHOD OBLIQUE REFLECTION PROBLEM IN A CONVEX POLYHEDRON

THE SKOROKHOD OBLIQUE REFLECTION PROBLEM IN A CONVEX POLYHEDRON GEORGIAN MATHEMATICAL JOURNAL: Vol. 3, No. 2, 1996, 153-176 THE SKOROKHOD OBLIQUE REFLECTION PROBLEM IN A CONVEX POLYHEDRON M. SHASHIASHVILI Abstract. The Skorokhod oblique reflection problem is studied

More information

08a. Operators on Hilbert spaces. 1. Boundedness, continuity, operator norms

08a. Operators on Hilbert spaces. 1. Boundedness, continuity, operator norms (February 24, 2017) 08a. Operators on Hilbert spaces Paul Garrett garrett@math.umn.edu http://www.math.umn.edu/ garrett/ [This document is http://www.math.umn.edu/ garrett/m/real/notes 2016-17/08a-ops

More information

Memoirs of My Research on Stochastic Analysis

Memoirs of My Research on Stochastic Analysis Memoirs of My Research on Stochastic Analysis Kiyosi Itô Professor Emeritus, Kyoto University, Kyoto, 606-8501 Japan It is with great honor that I learned of the 2005 Oslo Symposium on Stochastic Analysis

More information

Notes 1 : Measure-theoretic foundations I

Notes 1 : Measure-theoretic foundations I Notes 1 : Measure-theoretic foundations I Math 733-734: Theory of Probability Lecturer: Sebastien Roch References: [Wil91, Section 1.0-1.8, 2.1-2.3, 3.1-3.11], [Fel68, Sections 7.2, 8.1, 9.6], [Dur10,

More information

Boundedly complete weak-cauchy basic sequences in Banach spaces with the PCP

Boundedly complete weak-cauchy basic sequences in Banach spaces with the PCP Journal of Functional Analysis 253 (2007) 772 781 www.elsevier.com/locate/jfa Note Boundedly complete weak-cauchy basic sequences in Banach spaces with the PCP Haskell Rosenthal Department of Mathematics,

More information

S. DUTTA AND T. S. S. R. K. RAO

S. DUTTA AND T. S. S. R. K. RAO ON WEAK -EXTREME POINTS IN BANACH SPACES S. DUTTA AND T. S. S. R. K. RAO Abstract. We study the extreme points of the unit ball of a Banach space that remain extreme when considered, under canonical embedding,

More information

PROBABILITY: LIMIT THEOREMS II, SPRING HOMEWORK PROBLEMS

PROBABILITY: LIMIT THEOREMS II, SPRING HOMEWORK PROBLEMS PROBABILITY: LIMIT THEOREMS II, SPRING 15. HOMEWORK PROBLEMS PROF. YURI BAKHTIN Instructions. You are allowed to work on solutions in groups, but you are required to write up solutions on your own. Please

More information

A DECOMPOSITION THEOREM FOR FRAMES AND THE FEICHTINGER CONJECTURE

A DECOMPOSITION THEOREM FOR FRAMES AND THE FEICHTINGER CONJECTURE PROCEEDINGS OF THE AMERICAN MATHEMATICAL SOCIETY Volume 00, Number 0, Pages 000 000 S 0002-9939(XX)0000-0 A DECOMPOSITION THEOREM FOR FRAMES AND THE FEICHTINGER CONJECTURE PETER G. CASAZZA, GITTA KUTYNIOK,

More information

WAVELET EXPANSIONS OF DISTRIBUTIONS

WAVELET EXPANSIONS OF DISTRIBUTIONS WAVELET EXPANSIONS OF DISTRIBUTIONS JASSON VINDAS Abstract. These are lecture notes of a talk at the School of Mathematics of the National University of Costa Rica. The aim is to present a wavelet expansion

More information

Contents. 1 Preliminaries 3. Martingales

Contents. 1 Preliminaries 3. Martingales Table of Preface PART I THE FUNDAMENTAL PRINCIPLES page xv 1 Preliminaries 3 2 Martingales 9 2.1 Martingales and examples 9 2.2 Stopping times 12 2.3 The maximum inequality 13 2.4 Doob s inequality 14

More information

Spatial Ergodicity of the Harris Flows

Spatial Ergodicity of the Harris Flows Communications on Stochastic Analysis Volume 11 Number 2 Article 6 6-217 Spatial Ergodicity of the Harris Flows E.V. Glinyanaya Institute of Mathematics NAS of Ukraine, glinkate@gmail.com Follow this and

More information

PREDICTABLE REPRESENTATION PROPERTY OF SOME HILBERTIAN MARTINGALES. 1. Introduction.

PREDICTABLE REPRESENTATION PROPERTY OF SOME HILBERTIAN MARTINGALES. 1. Introduction. Acta Math. Univ. Comenianae Vol. LXXVII, 1(28), pp. 123 128 123 PREDICTABLE REPRESENTATION PROPERTY OF SOME HILBERTIAN MARTINGALES M. EL KADIRI Abstract. We prove as for the real case that a martingale

More information

Summable processes which are not semimartingales

Summable processes which are not semimartingales Summable processes which are not semimartingales Nicolae Dinculeanu Oana Mocioalca University of Florida, Gainesville, FL 32611 Abstract The classical stochastic integral HdX is defined for real-valued

More information

ALMOST AUTOMORPHIC GENERALIZED FUNCTIONS

ALMOST AUTOMORPHIC GENERALIZED FUNCTIONS Novi Sad J. Math. Vol. 45 No. 1 2015 207-214 ALMOST AUTOMOPHIC GENEALIZED FUNCTIONS Chikh Bouzar 1 Mohammed Taha Khalladi 2 and Fatima Zohra Tchouar 3 Abstract. The paper deals with a new algebra of generalized

More information

Vector measures of bounded γ-variation and stochastic integrals

Vector measures of bounded γ-variation and stochastic integrals Vector measures of bounded γ-variation and stochastic integrals Jan van Neerven and Lutz Weis bstract. We introduce the class of vector measures of bounded γ-variation and study its relationship with vector-valued

More information

ON THE REGULARITY OF ONE PARAMETER TRANSFORMATION GROUPS IN BARRELED LOCALLY CONVEX TOPOLOGICAL VECTOR SPACES

ON THE REGULARITY OF ONE PARAMETER TRANSFORMATION GROUPS IN BARRELED LOCALLY CONVEX TOPOLOGICAL VECTOR SPACES Communications on Stochastic Analysis Vol. 9, No. 3 (2015) 413-418 Serials Publications www.serialspublications.com ON THE REGULARITY OF ONE PARAMETER TRANSFORMATION GROUPS IN BARRELED LOCALLY CONVEX TOPOLOGICAL

More information

Ann. Funct. Anal. 1 (2010), no. 1, A nnals of F unctional A nalysis ISSN: (electronic) URL:

Ann. Funct. Anal. 1 (2010), no. 1, A nnals of F unctional A nalysis ISSN: (electronic) URL: Ann. Funct. Anal. (00), no., 44 50 A nnals of F unctional A nalysis ISSN: 008-875 (electronic) URL: www.emis.de/journals/afa/ A FIXED POINT APPROACH TO THE STABILITY OF ϕ-morphisms ON HILBERT C -MODULES

More information

ON PARABOLIC HARNACK INEQUALITY

ON PARABOLIC HARNACK INEQUALITY ON PARABOLIC HARNACK INEQUALITY JIAXIN HU Abstract. We show that the parabolic Harnack inequality is equivalent to the near-diagonal lower bound of the Dirichlet heat kernel on any ball in a metric measure-energy

More information

Kernel Method: Data Analysis with Positive Definite Kernels

Kernel Method: Data Analysis with Positive Definite Kernels Kernel Method: Data Analysis with Positive Definite Kernels 2. Positive Definite Kernel and Reproducing Kernel Hilbert Space Kenji Fukumizu The Institute of Statistical Mathematics. Graduate University

More information

Some SDEs with distributional drift Part I : General calculus. Flandoli, Franco; Russo, Francesco; Wolf, Jochen

Some SDEs with distributional drift Part I : General calculus. Flandoli, Franco; Russo, Francesco; Wolf, Jochen Title Author(s) Some SDEs with distributional drift Part I : General calculus Flandoli, Franco; Russo, Francesco; Wolf, Jochen Citation Osaka Journal of Mathematics. 4() P.493-P.54 Issue Date 3-6 Text

More information

PACKING-DIMENSION PROFILES AND FRACTIONAL BROWNIAN MOTION

PACKING-DIMENSION PROFILES AND FRACTIONAL BROWNIAN MOTION PACKING-DIMENSION PROFILES AND FRACTIONAL BROWNIAN MOTION DAVAR KHOSHNEVISAN AND YIMIN XIAO Abstract. In order to compute the packing dimension of orthogonal projections Falconer and Howroyd 997) introduced

More information

MATHS 730 FC Lecture Notes March 5, Introduction

MATHS 730 FC Lecture Notes March 5, Introduction 1 INTRODUCTION MATHS 730 FC Lecture Notes March 5, 2014 1 Introduction Definition. If A, B are sets and there exists a bijection A B, they have the same cardinality, which we write as A, #A. If there exists

More information

MATH MEASURE THEORY AND FOURIER ANALYSIS. Contents

MATH MEASURE THEORY AND FOURIER ANALYSIS. Contents MATH 3969 - MEASURE THEORY AND FOURIER ANALYSIS ANDREW TULLOCH Contents 1. Measure Theory 2 1.1. Properties of Measures 3 1.2. Constructing σ-algebras and measures 3 1.3. Properties of the Lebesgue measure

More information

TWO MAPPINGS RELATED TO SEMI-INNER PRODUCTS AND THEIR APPLICATIONS IN GEOMETRY OF NORMED LINEAR SPACES. S.S. Dragomir and J.J.

TWO MAPPINGS RELATED TO SEMI-INNER PRODUCTS AND THEIR APPLICATIONS IN GEOMETRY OF NORMED LINEAR SPACES. S.S. Dragomir and J.J. RGMIA Research Report Collection, Vol. 2, No. 1, 1999 http://sci.vu.edu.au/ rgmia TWO MAPPINGS RELATED TO SEMI-INNER PRODUCTS AND THEIR APPLICATIONS IN GEOMETRY OF NORMED LINEAR SPACES S.S. Dragomir and

More information

Measurable Choice Functions

Measurable Choice Functions (January 19, 2013) Measurable Choice Functions Paul Garrett garrett@math.umn.edu http://www.math.umn.edu/ garrett/ [This document is http://www.math.umn.edu/ garrett/m/fun/choice functions.pdf] This note

More information

Convergence of Banach valued stochastic processes of Pettis and McShane integrable functions

Convergence of Banach valued stochastic processes of Pettis and McShane integrable functions Convergence of Banach valued stochastic processes of Pettis and McShane integrable functions V. Marraffa Department of Mathematics, Via rchirafi 34, 90123 Palermo, Italy bstract It is shown that if (X

More information

1 Math 241A-B Homework Problem List for F2015 and W2016

1 Math 241A-B Homework Problem List for F2015 and W2016 1 Math 241A-B Homework Problem List for F2015 W2016 1.1 Homework 1. Due Wednesday, October 7, 2015 Notation 1.1 Let U be any set, g be a positive function on U, Y be a normed space. For any f : U Y let

More information

Filtrations, Markov Processes and Martingales. Lectures on Lévy Processes and Stochastic Calculus, Braunschweig, Lecture 3: The Lévy-Itô Decomposition

Filtrations, Markov Processes and Martingales. Lectures on Lévy Processes and Stochastic Calculus, Braunschweig, Lecture 3: The Lévy-Itô Decomposition Filtrations, Markov Processes and Martingales Lectures on Lévy Processes and Stochastic Calculus, Braunschweig, Lecture 3: The Lévy-Itô Decomposition David pplebaum Probability and Statistics Department,

More information

Ergodic Theorems. Samy Tindel. Purdue University. Probability Theory 2 - MA 539. Taken from Probability: Theory and examples by R.

Ergodic Theorems. Samy Tindel. Purdue University. Probability Theory 2 - MA 539. Taken from Probability: Theory and examples by R. Ergodic Theorems Samy Tindel Purdue University Probability Theory 2 - MA 539 Taken from Probability: Theory and examples by R. Durrett Samy T. Ergodic theorems Probability Theory 1 / 92 Outline 1 Definitions

More information

The Lévy-Itô decomposition and the Lévy-Khintchine formula in31 themarch dual of 2014 a nuclear 1 space. / 20

The Lévy-Itô decomposition and the Lévy-Khintchine formula in31 themarch dual of 2014 a nuclear 1 space. / 20 The Lévy-Itô decomposition and the Lévy-Khintchine formula in the dual of a nuclear space. Christian Fonseca-Mora School of Mathematics and Statistics, University of Sheffield, UK Talk at "Stochastic Processes

More information

An almost sure invariance principle for additive functionals of Markov chains

An almost sure invariance principle for additive functionals of Markov chains Statistics and Probability Letters 78 2008 854 860 www.elsevier.com/locate/stapro An almost sure invariance principle for additive functionals of Markov chains F. Rassoul-Agha a, T. Seppäläinen b, a Department

More information

Packing-Dimension Profiles and Fractional Brownian Motion

Packing-Dimension Profiles and Fractional Brownian Motion Under consideration for publication in Math. Proc. Camb. Phil. Soc. 1 Packing-Dimension Profiles and Fractional Brownian Motion By DAVAR KHOSHNEVISAN Department of Mathematics, 155 S. 1400 E., JWB 233,

More information

Brownian Motion and Conditional Probability

Brownian Motion and Conditional Probability Math 561: Theory of Probability (Spring 2018) Week 10 Brownian Motion and Conditional Probability 10.1 Standard Brownian Motion (SBM) Brownian motion is a stochastic process with both practical and theoretical

More information

On pathwise stochastic integration

On pathwise stochastic integration On pathwise stochastic integration Rafa l Marcin Lochowski Afican Institute for Mathematical Sciences, Warsaw School of Economics UWC seminar Rafa l Marcin Lochowski (AIMS, WSE) On pathwise stochastic

More information

On Least Squares Linear Regression Without Second Moment

On Least Squares Linear Regression Without Second Moment On Least Squares Linear Regression Without Second Moment BY RAJESHWARI MAJUMDAR University of Connecticut If \ and ] are real valued random variables such that the first moments of \, ], and \] exist and

More information

Recall that if X is a compact metric space, C(X), the space of continuous (real-valued) functions on X, is a Banach space with the norm

Recall that if X is a compact metric space, C(X), the space of continuous (real-valued) functions on X, is a Banach space with the norm Chapter 13 Radon Measures Recall that if X is a compact metric space, C(X), the space of continuous (real-valued) functions on X, is a Banach space with the norm (13.1) f = sup x X f(x). We want to identify

More information

WEYL S LEMMA, ONE OF MANY. Daniel W. Stroock

WEYL S LEMMA, ONE OF MANY. Daniel W. Stroock WEYL S LEMMA, ONE OF MANY Daniel W Stroock Abstract This note is a brief, and somewhat biased, account of the evolution of what people working in PDE s call Weyl s Lemma about the regularity of solutions

More information

Convergence rate estimates for the gradient differential inclusion

Convergence rate estimates for the gradient differential inclusion Convergence rate estimates for the gradient differential inclusion Osman Güler November 23 Abstract Let f : H R { } be a proper, lower semi continuous, convex function in a Hilbert space H. The gradient

More information

MATH 205C: STATIONARY PHASE LEMMA

MATH 205C: STATIONARY PHASE LEMMA MATH 205C: STATIONARY PHASE LEMMA For ω, consider an integral of the form I(ω) = e iωf(x) u(x) dx, where u Cc (R n ) complex valued, with support in a compact set K, and f C (R n ) real valued. Thus, I(ω)

More information

PROBABILITY: LIMIT THEOREMS II, SPRING HOMEWORK PROBLEMS

PROBABILITY: LIMIT THEOREMS II, SPRING HOMEWORK PROBLEMS PROBABILITY: LIMIT THEOREMS II, SPRING 218. HOMEWORK PROBLEMS PROF. YURI BAKHTIN Instructions. You are allowed to work on solutions in groups, but you are required to write up solutions on your own. Please

More information

I teach myself... Hilbert spaces

I teach myself... Hilbert spaces I teach myself... Hilbert spaces by F.J.Sayas, for MATH 806 November 4, 2015 This document will be growing with the semester. Every in red is for you to justify. Even if we start with the basic definition

More information

FURTHER STUDIES OF STRONGLY AMENABLE -REPRESENTATIONS OF LAU -ALGEBRAS

FURTHER STUDIES OF STRONGLY AMENABLE -REPRESENTATIONS OF LAU -ALGEBRAS FURTHER STUDIES OF STRONGLY AMENABLE -REPRESENTATIONS OF LAU -ALGEBRAS FATEMEH AKHTARI and RASOUL NASR-ISFAHANI Communicated by Dan Timotin The new notion of strong amenability for a -representation of

More information

Topological vectorspaces

Topological vectorspaces (July 25, 2011) Topological vectorspaces Paul Garrett garrett@math.umn.edu http://www.math.umn.edu/ garrett/ Natural non-fréchet spaces Topological vector spaces Quotients and linear maps More topological

More information

The Azéma-Yor Embedding in Non-Singular Diffusions

The Azéma-Yor Embedding in Non-Singular Diffusions Stochastic Process. Appl. Vol. 96, No. 2, 2001, 305-312 Research Report No. 406, 1999, Dept. Theoret. Statist. Aarhus The Azéma-Yor Embedding in Non-Singular Diffusions J. L. Pedersen and G. Peskir Let

More information

b i (µ, x, s) ei ϕ(x) µ s (dx) ds (2) i=1

b i (µ, x, s) ei ϕ(x) µ s (dx) ds (2) i=1 NONLINEAR EVOLTION EQATIONS FOR MEASRES ON INFINITE DIMENSIONAL SPACES V.I. Bogachev 1, G. Da Prato 2, M. Röckner 3, S.V. Shaposhnikov 1 The goal of this work is to prove the existence of a solution to

More information

arxiv:math/ v1 [math.fa] 26 Oct 1993

arxiv:math/ v1 [math.fa] 26 Oct 1993 arxiv:math/9310217v1 [math.fa] 26 Oct 1993 ON COMPLEMENTED SUBSPACES OF SUMS AND PRODUCTS OF BANACH SPACES M.I.Ostrovskii Abstract. It is proved that there exist complemented subspaces of countable topological

More information

Finite pseudocomplemented lattices: The spectra and the Glivenko congruence

Finite pseudocomplemented lattices: The spectra and the Glivenko congruence Finite pseudocomplemented lattices: The spectra and the Glivenko congruence T. Katriňák and J. Guričan Abstract. Recently, Grätzer, Gunderson and Quackenbush have characterized the spectra of finite pseudocomplemented

More information

Doléans measures. Appendix C. C.1 Introduction

Doléans measures. Appendix C. C.1 Introduction Appendix C Doléans measures C.1 Introduction Once again all random processes will live on a fixed probability space (Ω, F, P equipped with a filtration {F t : 0 t 1}. We should probably assume the filtration

More information

n E(X t T n = lim X s Tn = X s

n E(X t T n = lim X s Tn = X s Stochastic Calculus Example sheet - Lent 15 Michael Tehranchi Problem 1. Let X be a local martingale. Prove that X is a uniformly integrable martingale if and only X is of class D. Solution 1. If If direction:

More information

On a Class of Multidimensional Optimal Transportation Problems

On a Class of Multidimensional Optimal Transportation Problems Journal of Convex Analysis Volume 10 (2003), No. 2, 517 529 On a Class of Multidimensional Optimal Transportation Problems G. Carlier Université Bordeaux 1, MAB, UMR CNRS 5466, France and Université Bordeaux

More information

L -uniqueness of Schrödinger operators on a Riemannian manifold

L -uniqueness of Schrödinger operators on a Riemannian manifold L -uniqueness of Schrödinger operators on a Riemannian manifold Ludovic Dan Lemle Abstract. The main purpose of this paper is to study L -uniqueness of Schrödinger operators and generalized Schrödinger

More information

Maximum Process Problems in Optimal Control Theory

Maximum Process Problems in Optimal Control Theory J. Appl. Math. Stochastic Anal. Vol. 25, No., 25, (77-88) Research Report No. 423, 2, Dept. Theoret. Statist. Aarhus (2 pp) Maximum Process Problems in Optimal Control Theory GORAN PESKIR 3 Given a standard

More information

G(t) := i. G(t) = 1 + e λut (1) u=2

G(t) := i. G(t) = 1 + e λut (1) u=2 Note: a conjectured compactification of some finite reversible MCs There are two established theories which concern different aspects of the behavior of finite state Markov chains as the size of the state

More information

{σ x >t}p x. (σ x >t)=e at.

{σ x >t}p x. (σ x >t)=e at. 3.11. EXERCISES 121 3.11 Exercises Exercise 3.1 Consider the Ornstein Uhlenbeck process in example 3.1.7(B). Show that the defined process is a Markov process which converges in distribution to an N(0,σ

More information

THE L 2 -HODGE THEORY AND REPRESENTATION ON R n

THE L 2 -HODGE THEORY AND REPRESENTATION ON R n THE L 2 -HODGE THEORY AND REPRESENTATION ON R n BAISHENG YAN Abstract. We present an elementary L 2 -Hodge theory on whole R n based on the minimization principle of the calculus of variations and some

More information

n [ F (b j ) F (a j ) ], n j=1(a j, b j ] E (4.1)

n [ F (b j ) F (a j ) ], n j=1(a j, b j ] E (4.1) 1.4. CONSTRUCTION OF LEBESGUE-STIELTJES MEASURES In this section we shall put to use the Carathéodory-Hahn theory, in order to construct measures with certain desirable properties first on the real line

More information

An Infinitesimal Approach to Stochastic Analysis on Abstract Wiener Spaces

An Infinitesimal Approach to Stochastic Analysis on Abstract Wiener Spaces An Infinitesimal Approach to Stochastic Analysis on Abstract Wiener Spaces Dissertation zur Erlangung des akademischen Grades eines Doktors der Naturwissenschaften an der Fakultät für Mathematik, Informatik

More information

Chapter One. The Calderón-Zygmund Theory I: Ellipticity

Chapter One. The Calderón-Zygmund Theory I: Ellipticity Chapter One The Calderón-Zygmund Theory I: Ellipticity Our story begins with a classical situation: convolution with homogeneous, Calderón- Zygmund ( kernels on R n. Let S n 1 R n denote the unit sphere

More information

Fréchet algebras of finite type

Fréchet algebras of finite type Fréchet algebras of finite type MK Kopp Abstract The main objects of study in this paper are Fréchet algebras having an Arens Michael representation in which every Banach algebra is finite dimensional.

More information

Lecture 22 Girsanov s Theorem

Lecture 22 Girsanov s Theorem Lecture 22: Girsanov s Theorem of 8 Course: Theory of Probability II Term: Spring 25 Instructor: Gordan Zitkovic Lecture 22 Girsanov s Theorem An example Consider a finite Gaussian random walk X n = n

More information

STAT 331. Martingale Central Limit Theorem and Related Results

STAT 331. Martingale Central Limit Theorem and Related Results STAT 331 Martingale Central Limit Theorem and Related Results In this unit we discuss a version of the martingale central limit theorem, which states that under certain conditions, a sum of orthogonal

More information

16.1. Signal Process Observation Process The Filtering Problem Change of Measure

16.1. Signal Process Observation Process The Filtering Problem Change of Measure 1. Introduction The following notes aim to provide a very informal introduction to Stochastic Calculus, and especially to the Itô integral. They owe a great deal to Dan Crisan s Stochastic Calculus and

More information

Additive Summable Processes and their Stochastic Integral

Additive Summable Processes and their Stochastic Integral Additive Summable Processes and their Stochastic Integral August 19, 2005 Abstract We define and study a class of summable processes,called additive summable processes, that is larger than the class used

More information

at time t, in dimension d. The index i varies in a countable set I. We call configuration the family, denoted generically by Φ: U (x i (t) x j (t))

at time t, in dimension d. The index i varies in a countable set I. We call configuration the family, denoted generically by Φ: U (x i (t) x j (t)) Notations In this chapter we investigate infinite systems of interacting particles subject to Newtonian dynamics Each particle is characterized by its position an velocity x i t, v i t R d R d at time

More information

3 (Due ). Let A X consist of points (x, y) such that either x or y is a rational number. Is A measurable? What is its Lebesgue measure?

3 (Due ). Let A X consist of points (x, y) such that either x or y is a rational number. Is A measurable? What is its Lebesgue measure? MA 645-4A (Real Analysis), Dr. Chernov Homework assignment 1 (Due ). Show that the open disk x 2 + y 2 < 1 is a countable union of planar elementary sets. Show that the closed disk x 2 + y 2 1 is a countable

More information

Wiener Measure and Brownian Motion

Wiener Measure and Brownian Motion Chapter 16 Wiener Measure and Brownian Motion Diffusion of particles is a product of their apparently random motion. The density u(t, x) of diffusing particles satisfies the diffusion equation (16.1) u

More information

Patryk Pagacz. Characterization of strong stability of power-bounded operators. Uniwersytet Jagielloński

Patryk Pagacz. Characterization of strong stability of power-bounded operators. Uniwersytet Jagielloński Patryk Pagacz Uniwersytet Jagielloński Characterization of strong stability of power-bounded operators Praca semestralna nr 3 (semestr zimowy 2011/12) Opiekun pracy: Jaroslav Zemanek CHARACTERIZATION OF

More information

ON COMPLEMENTED SUBSPACES OF SUMS AND PRODUCTS OF BANACH SPACES

ON COMPLEMENTED SUBSPACES OF SUMS AND PRODUCTS OF BANACH SPACES PROCEEDINGS OF THE AMERICAN MATHEMATICAL SOCIETY Volume 124, Number 7, July 1996 ON COMPLEMENTED SUBSPACES OF SUMS AND PRODUCTS OF BANACH SPACES M. I. OSTROVSKII (Communicated by Dale Alspach) Abstract.

More information

THE DUAL FORM OF THE APPROXIMATION PROPERTY FOR A BANACH SPACE AND A SUBSPACE. In memory of A. Pe lczyński

THE DUAL FORM OF THE APPROXIMATION PROPERTY FOR A BANACH SPACE AND A SUBSPACE. In memory of A. Pe lczyński THE DUAL FORM OF THE APPROXIMATION PROPERTY FOR A BANACH SPACE AND A SUBSPACE T. FIGIEL AND W. B. JOHNSON Abstract. Given a Banach space X and a subspace Y, the pair (X, Y ) is said to have the approximation

More information