Assistant Professor in Probability Theory & Stochastic Analysis


📍 Aarhus University – Department of Mathematics (Ny Munkegade 118, 8000 Aarhus C)
🗓 Position start: 1 August 2026 (3-year full-time fixed-term: 1 Aug 2026–31 Jul 2029)

Join the vibrant Stochastics Group at the Department of Mathematics, Aarhus University, at the faculty of Natural Sciences. Our research spans probability theory, stochastic analysis, stochastic differential geometry, Malliavin calculus, stochastic PDEs, random matrices, and related areas of analysis and geometry. The successful candidate will be affiliated with the “Stochastic Analysis in Aarhus” project, funded by the Villum Foundation and led by myself.

What you’ll bring:
A completed PhD in mathematics, with a strong research record in probability theory, stochastic analysis or closely related fields.
Demonstrated ability to conduct independent research and to collaborate as part of a research group.
Experience in teaching at university level and willingness to contribute to research-based teaching, including examination duties.

What we offer:
A stimulating research environment with an international profile and strong collaborations in stochastic analysis and related fields.
Excellent research infrastructure, including access to travel funds and visitor support.
A collegial, dynamic department with active seminar series, workshops and a broad range of research groups.
Support for international researchers and accompanying families, including relocation services and career counselling for expat partners.

Application deadline: 15 December 2025 – 23:59 CET

If you’re passionate about deepening your impact in stochastic analysis in a supportive, internationally engaged academic setting, we would love to hear from you.

💡 Feel free to share this opportunity with suitable colleagues or post-docs.

https://math.au.dk/om/ledige-stillinger/job/assistant-professor-in-probability-theory-and-stochastic-analysis

Posted in Uncategorized | Leave a comment

Aarhus Summer School in Analysis and Probability

The 6–10 July, 2026, Aarhus Summer School in Analysis and Probability aims to bring together leading experts and young researchers in the areas of analysis, probability theory, and their interactions. The school will feature a series of five hours lecture courses delivered by distinguished speakers, offering participants the opportunity to engage with current developments at the forefront of the field. In addition to the main lectures, the program will include short talks by graduate students and early-career researchers. The summer school will  provide a stimulating and friendly environment, ideal for fostering discussions, collaborations, and the exchange of ideas among researchers at all career stages.

Funding for Early Career Researchers

We are pleased to announce that limited funding will be available for early career researchers (PhD students and postdocs) who wish to participate in the summer school and propose to give a talk. Applicants interested in being considered for funding are invited to include a short abstract of their proposed presentation when registering or submitting their application to the organisers.

Registration will begin on December 1st.

https://conferences.au.dk/saa-summer-school-2026

The courses are the following.

Energy measures and applications by Mathav Murugan

I plan to cover some results on energy measures such as singularity/absolute continuity, its role in the attainment problem for conformal walk dimension (and Ahlfors regular conformal dimension), and also applications to martingale dimension.

The Sard conjecture in sub-Riemannian geometry by Luca Rizzi 

Sard’s theorem asserts that the set of critical values of a smooth map between finite-dimensional manifolds has measure zero. It is well-known, however, that when the domain is infinite dimensional and the range is finite dimensional, the result is not true. Counterexamples are known already in the class of polynomial maps from an infinite-dimensional Hilbert space to the Euclidean line. The problem is in particular relevant in sub-Riemannian geometry, where the validity of the above property for the endpoint map is the so-called Sard conjecture, and is one of the main open problems in the field. We present an overview of the Sard problem in sub-Riemannian geometry together with some recent results in collaboration with Lerario and Tiberio.

Gromov hyperbolicity, uniformization, and potential theory by Nageswari Shanmugalingam

In the non smooth setting of metric measure spaces, where related notions of interest are invariant under bi-Lipschitz transformations of a metric space, negative curvature has a counterpart as Gromov hyperbolicity. This is a coarser, more large-scale notion, but appears to be closely associated with bounded uniform domains via a transformation called uniformization. In this series of lectures we will discuss how such a transformation can link potential theory on Gromov hyperbolic spaces with potential theory on bounded uniform domains. We will begin the course by discussing analogs of Sobolev spaces in non-smooth setting.

Some aspects of parabolic Anderson models by Samy Tindel 

 In this mini-course I will introduce some basic elements allowing to understand parabolic Anderson models on geometric structures. The topics covered will be:

(1) Gaussian noises and their regularity.

(2) Existence and uniqueness results for the stochastic heat equation.

(3) Moment estimates.

(4) Extensions from Euclidian spaces to hypoelliptic settings, fractals and other geometric contexts.

Notice that I’m not assuming any knowledge of stochastic analysis from the audience. I will try to introduce the objects I’m manipulating in a self-contained way.

Posted in Uncategorized | Tagged , , , , , | Leave a comment

Postdoctoral Positions in Probability Theory and Analysis

Several postdoctoral positions at Aarhus are opening in my group. The earliest possible start date is February 1st 2026, with flexibility for later starts in the spring.

These positions are supported by a Villum Investigator Grant and the ERC, offering excellent opportunities for travels, research and international collaboration.

https://international.au.dk/about/profile/vacant-positions/job/postdoctoral-positions-in-probability-theory-and-analysis-aarhus-university

Posted in Uncategorized | Leave a comment

Lecture notes: A short introduction to Dirichlet spaces

Here are the lecture notes of the minicourse given at NYU Abu Dhabi.

Posted in Dirichlet forms at NYU, Lecture Notes | Tagged , , , , , | Leave a comment

Lecture 7. Further topics

In this lecture let X be a locally compact and complete metric space equipped with a Radon measure μ supported on X. Let (ℰ, ℱ = dom(ℰ)) be a Dirichlet form on X. We assume throughout that the heat semigroup Pt is stochastically complete, i.e., Pt1 = 1 for every t ≥ 0.

Contents

Regular Dirichlet forms, Energy measures

We denote by Cc(X) the vector space of all continuous functions with compact support in X and C0(X) its closure with respect to the supremum norm.

A core for (X,μ,ℰ,ℱ) is a subset 𝒞 of Cc(X) ∩ ℱ which is dense in Cc(X) in the supremum norm and dense in in the norm

(||f||^2_{L^2(X,\mu)} + \mathcal{E}(f,f))^{1/2}.

Definition. The Dirichlet form is called regular if it admits a core.

Recall that for any f,g ∈ ℱ, we have

\mathcal{E}(f,g)=\lim_{t\to 0}\frac1t\langle (I-P_t)f,g\rangle=\lim_{t\to 0}\frac1{t}\int_X \int_X(f(x)-f(y)) g(x) p_t(x,dy) d\mu(x),

where pt(x,·) are the heat kernel measures associated to the Dirichlet form (ℰ, ℱ).
From the symmetry property \eqref{heat kernel measure symmetry} of the heat kernel measure one also has

\mathcal{E}(f, g)=\lim_{t\to 0}\frac1{2t}\int_X \int_X(f(x)-f(y))(g(x)-g(y)) p_t(x,dy) d\mu(x).

Lemma. For f,g ∈ ℱ ∩ L(X,μ), fg ∈ ℱ and

\mathcal{E}(fg)^{1/2} \le ||f||_\infty \mathcal{E}(g)^{1/2} +||g||_\infty \mathcal{E}(f)^{1/2}.

Proof. For f, g ∈ ℱ ∩ L(X,μ) such that fg ∈ ℱ,

\mathcal{E}(fg)=\lim_{t\to 0}\frac1{2t}\int_X \int_X(f(x)g(x)-f(y)g(y))^2 p_t(x,dy) d\mu(x).

Write f(x)g(x) – f(y)g(y) = f(x)(g(x) – g(y)) + g(y)(f(x) – f(y)), then by Minkowski’s inequality

\left(\int_X \int_X(f(x)g(x)-f(y)g(y))^2 p_t(x,dy) d\mu(x)\right)^{1/2}

\le ||f||_{\infty}\left(\int_X \int_X (g(x)-g(y))^2 p_t(x,dy) d\mu(x)\right)^{1/2}

+||g||_{\infty}\left(\int_X \int_X(f(x)-f(y))^2 p_t(x,dy) d\mu(x)\right)^{1/2}.

We conclude the expected inequality by multiplying by 1/√2t and taking the limit t → 0 for both sides above.

Theorem (Energy measures) Assume that is regular. For f ∈ ℱ ∩ L(X,μ), there exists a unique Radon measure on X, denoted by dΓ(f), so that for every ϕ ∈ ℱ ∩ Cc(X),

\int_X\phi\, d\Gamma(f)=\frac{1}{2}[2\mathcal{E}(\phi f,f)-\mathcal{E}(\phi, f^2)]

= \lim_{t \to 0} \frac{1}{2t} \int_X \int_X \phi(x) (f(x)-f(y))^2 p_t(x,dy) d\mu(x).

The Radon measure dΓ(f) is called the energy measure of f (and is therefore the weak * limit of (1/(2t)) ∫X (f(x) – f(y))2 pt(x,dy)).

Proof. Let f ∈ ℱ ∩ L(X,μ). For any ϕ ∈ ℱ ∩ Cc(X),

\frac1{2t} \int_X \int_X \phi(x) (f(x)-f(y))^2 p_t(x,dy) d\mu(x)=-\frac1{2t} \langle (I-P_t)f^2 ,\phi\rangle+\frac1{t} \langle (I-P_t)f ,f\phi\rangle.

Letting t → 0, the right-hand side converges to

\frac{1}{2}[2\mathcal{E}(\phi f,f)-\mathcal{E}(\phi, f^2)] .

On the other hand, observing that

\frac1{2t} \int_X \int_X |\phi(x)| (f(x)-f(y))^2 p_t(x,dy) d\mu(x) \le \|\phi\|_{\infty} \frac1{2t} \int_X \int_X (f(x)-f(y))^2 p_t(x,dy) d\mu(x),

we deduce

\left|\frac{1}{2}[2\mathcal{E}(\phi f,f)-\mathcal{E}(\phi, f^2)]\right|\le \|\phi\|_{\infty}\mathcal E(f).

Therefore we conclude the proof by applying the Riesz-Markov representation theorem.

One can actually define dΓ(f,f) for every f ∈ ℱ using the following lemmas.

Lemma. Let f ∈ ℱ. Then fn = min(n, max(-n, f)) ∈ ℱ and ℰ(f – fn) → 0.

Proof. Let f ∈ ℱ. For every x, y ∈ X, we have

|f_n(x) - f_n(y)| \leq |f(x) - f(y)|

and |fn(x)| ≤ |f(x)|. So fn is a normal contraction of f and ℰ(fn) ≤ ℰ(f).

Recall that

\mathcal{E}(f-f_n)=\lim_{t\to 0}\frac1{2t} \int_X \int_X(f(x)-f_n(x)-f(y)+f_n(y))^2 p_t(x,dy) d\mu(x).

Expanding the square inside the integral gives that

\mathcal{E}(f-f_n)=\mathcal E(f)+\mathcal E(f_n)-2\mathcal E(f_n,f)\le 2\mathcal E(f)-2\mathcal E(f_n,f).

Letting n → ∞, we have ℰ(fn,f) → ℰ(f). Therefore ℰ(f – fn) converges to 0 as n → ∞.

Lemma. For f,g ∈ ℱ ∩ L(X,μ) and nonnegative ϕ ∈ ℱ ∩ Cc(X),

\left| \sqrt{\int_X\phi\, d\Gamma(f)} -\sqrt{\int_X\phi\, d\Gamma(g) } \right|^2 \le \int_X\phi\, d\Gamma(f-g) \le ||\phi||_{L^\infty(X,\mu)}\mathcal{E}(f-g)

Proof. The second inequality follows from the proof in Theorem 16.3. For the first inequality, it suffice to show that for any f, g ∈ ℱ ∩ L(X,μ) and any nonnegative ϕ ∈ ℱ ∩ Cc(X),

\sqrt{\int_X\phi\, d\Gamma(f)} \le \sqrt{\int_X\phi\, d\Gamma(g) } +\sqrt{ \int_X\phi\, d\Gamma(f-g) }.

Indeed, this inequality follows from a similar proof as in Lemma 16.2 by noting that f = f – g + g and using Minkowski’s inequality.

Thanks to the previous lemmas, by approximation, one can define dΓ(f) for every f ∈ ℱ by

dΓ(f) = sup{dΓ(fn) : fn = min(n, max(-n, f)), n = 1, 2, …}.

For f,g ∈ ℱ, one can define dΓ(f,g) by polarization

dΓ(f,g) = (1/4)(dΓ(f + g) – dΓ(f – g)).

The following representation theorem for regular Dirichlet forms then holds.

Theorem  (Beurling-Deny) Assume that is regular. For u,v ∈ ℱ, ℰ(u,v) = ∫X dΓ(u,v).

Hunt process associated with a regular Dirichlet form

Definition. A Hunt process with state space X is a family of stochastic process (Xt)t≥0 and probability measures (ℙx)x∈X defined on a measure space (Ω, ℱ), such that (Xt)t≥0 is adapted w.r.t. the right-continuous minimal completed admissible filtration (ℱt)t≥0, X0 = x, x-a.s. and the following hold:

  • (i) x → ℙx(Xt ∈ B) is measurable for all t > 0 and B ∈ ℬ(X),
  • (ii) X is a strong Markov process, i.e. for every stopping time T, XT is T-measurable and for every B ∈ ℬ(X)

    \mathbb{P}_x(X_{T+t} \in B | \mathcal{F}_T) = \mathbb{P}_{X_T}(X_t \in B) \quad \mathbb{P}_x\text{-a.s. on } \{T < \infty\},

  • (iii) X is right-continuous, i.e.

    \lim_{s \downarrow t} X_s = X_t, \quad \forall t \quad \mathbb{P}_x\text{-a.s.}

  • (iv) X is quasi left-continuous, i.e. for all stopping times T and (Tn)n such that Tn ↑ T a.s.

    \lim_{n \to \infty} X_{T_n} = X_T, \quad \mathbb{P}_x\text{-a.s. on } \{T < \infty\}.

Remark 

  1. Note that quasi left-continuity does not necessarily imply left-continuity, because the set

    A = \left\{ \lim_{n \to \infty} X_{s_n} = X_t \right\}

    might depend on the choice of sequence (sn)n, sn ↑ t.

  2. In more generality, one might consider situations where (ℙx(Xt ∈ ·))x,t are sub-probability measures (i.e. all the axioms of probability measures are satisfied but x(Xt ∈ Ω) ≤ 1). In that case we can perform a one-point compactification of X by introducing a cemetery state ∂ ∉ X and redefine x to be a probability measure on X ∪ {∂}.

The following theorem can then be proved, see the book [FOT], Theorem 4.2.1.

Theorem (Fukushima) Assume that is regular, then there exists a Hunt process ((Xt)t≥0, (ℙx)x∈X) such that for μ-a.e. x ∈ X, A ∈ ℬ(X) and t ≥ 0,

\mathbb{P}_x (X_t \in A)=p_t(x,A)

where pt(x, ·) are the heat kernel measures associated to the Dirichlet form (ℰ, 𝒟(ℰ)).

Intrinsic metric

Definition. The Dirichlet form is called strongly local if for any two functions f,g ∈ ℱ with compact supports such that f is constant in a neighborhood of the support of g, we have ℰ(f,g) = 0.

With respect to we can define the following intrinsic metric d on X by

d_{\mathcal{E}}(x,y)=\sup\{u(x)-u(y)\, :\, u\in\mathcal{F}\cap C_0(X)\text{ and } d\Gamma(u,u)\le d\mu\}.

Here the condition dΓ(u,u) ≤ dμ means that Γ(u,u) is absolutely continuous with respect to μ with Radon-Nikodym derivative bounded by 1.

The term “intrinsic metric” is potentially misleading because in general there is no reason why d is a metric on X (it could be infinite for a given pair of points x,y or zero for some distinct pair of points).

Definition. A strongly local regular Dirichlet space is called strictly local if d is a metric on X and the topology induced by d coincides with the topology on X.

Example (Riemannian manifolds) Let (M,g) be a complete n-dimensional Riemannian manifold with Riemannian volume measure μ. We consider the standard Dirichlet form on M, which is obtained by closing the bilinear form

\mathcal{E}(f,g)=\int_\mathbb{M} \langle \nabla f ,\nabla g \rangle d\mu, \quad f,g \in C_0^\infty(\mathbb M).

Then is a strictly local Dirichlet form such that

d_\mathcal{E} (x,y)=d_g(x,y).

Example (Carnot groups) Let G be a Carnot group with sub-Laplacian

L=\sum_{i=1}^d V_i^2

and Dirichlet form

\mathcal{E}(f)=\int_{\mathbb G} \sum_{i=1} (V_if)^2 d\mu.

Then is a strictly local Dirichlet form such that

d_\mathcal{E} (x,y)=d_{CC} (x,y)

where dCC is the so-called Carnot-Carathéodory distance which is defined as follows.

An absolutely continuous curve γ : [0,T] → G is said to be subunit for the operator L if for every smooth function f : G → ℝ we have

\left| \frac{d}{dt} f ( \gamma(t) ) \right| \le \sqrt{ (\Gamma f) (\gamma(t)) }.

We then define the subunit length of γ as s(γ) = T.

Given x,y ∈ G, we indicate with

S(x,y) = {γ : [0,T] → G | γ is subunit for Γ, γ(0) = x, γ(T) = y}.

It is a consequence of the Chow-Rashevskii theorem that

S(x,y) ≠ ∅, for every x, y ∈ G.

One defines then

d_{CC}(x,y) = \inf\{ \ell_s(\gamma) \mid  \gamma \in  S(x,y)\}

Example Consider on the Sierpinski gasket the standard Dirichlet form . Then is regular, but unless f is constant, for f ∈ ℱ, dΓ(f) is singular with respect to the Hausdorff measure μ, see [KM19]. As a consequence is not strictly local.

 Further reading

Far from being exhaustive we mention the following references for further reading:

  1. The book [FOT] is a standard comprehensive reference in the theory of Dirichlet forms and associated Hunt processes, see also [ChenFukushima].
  2. The book [GSC] shows how one can restrict Dirichlet forms to domains (abstract Dirichlet and Neumann boundary conditions).
  3. Parabolic regularity theory for the heat equation can be developed in the setting of abstract strictly local Dirichlet spaces, see [Saloff].
Posted in Dirichlet forms at NYU | Tagged , , | Leave a comment

Lecture 6. Gagliardo-Nirenberg inequalities in Dirichlet spaces

Let (X,μ,ℰ,ℱ) be a Dirichlet space and {Pt}t∈[0,∞) denote the associated Markovian semigroup. Throughout the lecture, we shall assume that P_t admits a measurable heat kernel pt(x,y) satisfying, for some C > 0 and β > 0,

p_{t}(x,y)\leq C t^{-\beta}

for μ × μ-a.e. (x,y) ∈ X × X, and for each t ∈ (0,+∞). We also assume stochastic completeness, which means that Pt1 = 1 for every t ≥ 0.

The goal of the lecture is to prove the Gagliardo-Nirenberg inequalities in that setting. The techniques we use come from a paper by Bakry-Coulhon-Ledoux-Saloff-Coste. For another approach to the Sobolev inequality, we refer to the post.

Contents

Preliminary lemmas

Lemma. For every f ∈ ℱ, t ≥ 0,

|| P_t f -f ||_{L^2(X,\mu)} \le C \sqrt{t} \mathcal{E}(f,f)^{1/2}.

Proof: Let Δ be the generator of the semigroup (Pt)t≥0. We use spectral theorem to represent Δ as a multiplier in some L2(Ω,ν) space:

U^{-1} A U g (x)=-\lambda(x) g(x),

so that

|| P_t f -f ||_{L^2(X,\mu)}^2=\int_{\Omega} \left( 1-e^{-t\lambda(x)}\right)^2(U^{-1} f)^2(x) d\nu(x)

\le C t \int_{\Omega} \lambda(x) (U^{-1} f)^2(x) d\nu(x)

= Ct \mathcal{E}(f,f).

Lemma. For every f ∈ ℱ,

\lim_{t \to 0} t^{-1}\int_X \int_X |f (x)-f(y)|^2 p_t(x,y) d\mu(x)d\mu(y) =2 \mathcal{E}(f,f).

Proof: Using Pt1 = 1 we have

\int_X \int_X |f (x)-f(y)|^2 p_t(x,y) d\mu(x)d\mu(y)

\int_X \int_X( f (x)^2-2f(x)f(y)+f(y)^2) p_t(x,y) d\mu(x)d\mu(y)

=2\int_X f(x)^2 d\mu(x)-2\int_X f(x)(P_t f) (x)d\mu(x)

=2 \int_X (f(x)-P_tf(x)) f(x) d\mu(x)

and we conclude thanks to a previous result.

Sobolev inequality

Lemma. Let 1 ≤ q < +∞. There exists a constant C > 0 such that for every f ∈ ℱ ∩ Lq(X,μ) and s ≥ 0,

\sup_{s \ge 0} s^{1 +\frac{q}{2\beta}} \mu \left( \{ x \in X\, :\, | f(x) | > s \} \right)^{\frac{1}{2}}\le C \mathcal{E}(f,f)^{1/2} || f ||_{L^q(X,\mu)}^{\frac{q}{2\beta}}.

Proof: Let f ∈ ℱ and denote

F(s)=\mu \left( \{ x \in X\, :\, | f(x) | > s \} \right).

We have then

F(s) \le \mu \left( \{ x \in X\, :\, | f(x) -P_t f (x) | > s/2 \} \right)+\mu \left( \{ x \in X\, :\, | P_t f (x) | > s/2 \} \right).

Now, from the heat kernel upper bound pt(x,y) ≤ C t, t > 0, one deduces, for g ∈ L1(X,μ), that

|P_t g (x) | \le Ct^{-\beta} \| g \|_{L^1(X,\mu)}.

Since Pt is a contraction in L(X,μ), by the Riesz-Thorin interpolation one obtains

| P_t f (x) | \le \frac{C^{1/q}}{t^{\beta /q}} \| f \|_{L^q(X,\mu)}.

Therefore, for

s= 2 \frac{C^{1/q}}{t^{\frac{\beta}{ q}}} \| f \|_{L^q(X,\mu)},

one has

\mu \left( \{ x \in X\, :\, | P_t f (x) | > s/2 \} \right)=0.

On the other hand, from previous  lemma,

\mu \left( \{ x \in X\, :\, | f(x) -P_t f (x) | > s/2 \} \right) \le C s^{-2} t \mathcal{E}(f,f).

We conclude that

F(s)^{1/2} \le C s^{-1 -\frac{q}{2\beta}} \mathcal{E}(f,f)^{1/2} \| f \|_{L^q(X,\mu)}^{\frac{q}{2\beta}}.

Lemma. Assume β > 1. There exists a constant C > 0 such that for every f ∈ ℱ,

\sup_{s \ge 0}\, s\, \mu \left( \{ x \in X\, :\, | f(x) | \ge s \} \right)^{\frac{1}{q}} \le C \mathcal{E}(f,f)^{1/2},

where q = 2β/(β – 1).

Proof: Let f ∈ ℱ be a non-negative function. For k ∈ ℤ, we denote

f_k=(f-2^k)_+ \wedge 2^k.

Observe that fk ∈ L2(X,μ) and ||fk||L2(X,μ) ≤ ||f||L2(X,μ). Moreover, for every x, y ∈ X, |fk(x) – fk(y)| ≤ |f(x) – f(y)| and so ℰ(fk,fk) ≤ ℰ(f,f). We also note that fk ∈ L1(X,μ), with

||f_k||_{L^1(X,\mu)} =\int_X |f_k| d\mu \le 2^k \mu (\{ x \in X\, :\, f(x) \ge 2^k \}).

We now use the previous lemma to deduce:

\sup_{s \ge 0} s^{1 +\frac{1}{2\beta}} \mu \left( \{ x \in X\, :\, f_k(x) > s \} \right)^{\frac{1}{2}} \le C \mathcal{E}(f_k,f_k)^{1/2} || f_k \|_{L^1(X,\mu)}^{\frac{1}{2\beta}}

\le C \mathcal{E}(f_k,f_k)^{1/2} \left( 2^k \mu (\{ x \in X\, :\, f(x) \ge 2^k \})\right)^{\frac{1}{2\beta}}

In particular, by choosing s = 2k we obtain

2^{k \left(1 +\frac{1}{2\beta} \right)} \mu \left( \{ x \in X\, :\, f(x) \ge 2^{k+1} \} \right)^{\frac{1}{2}} \le C \mathcal{E}(f_k,f_k)^{1/2} \left( 2^k \mu (\{ x \in X\, :\, f(x) \ge 2^k \})\right)^{\frac{1}{2\beta}}

Let

M(f)=\sup_{k \in \mathbb{Z}} 2^k \mu (\{ x \in X\, :\, f(x) \ge 2^k \})^{1/q}

where q = 2β/(β – 1). Using the fact that 1/q = 1/2 – 1/(2β) and the previous inequality we obtain:

2^{k} \mu \left( \{ x \in X\, :\, f(x) \ge 2^{k+1} \} \right)^{\frac{1}{2}} \le C 2^{ -\frac{kq}{2\beta}}\mathcal{E}(f,f)^{1/2} M(f)^{\frac{q}{2\beta}}

and

2^k \mu \left( \{ x \in X\, :\, f(x) \ge 2^{k+1} \} \right)^{\frac{1}{q}} \le C^{\frac{2}{q} } \mathcal{E}(f,f)^{1/q}M(f)^{\frac{1}{\beta}}.

Therefore

M(f)^{1-\frac{1}{\beta}} \le 2 C^{\frac{2}{q} } \mathcal{E}(f,f)^{1/q}

and one concludes

M(f) \le 2^{q/2} C \mathcal{E}(f,f)^{1/2}.

This easily yields:

\sup_{s \ge 0} s \mu \left( \{ x \in X\, :\, f(x) \ge s \} \right)^{\frac{1}{q}} \le 2^{1+q/2} C \mathcal{E}(f,f)^{1/2}

Let now f ∈ ℱ, which is not necessarily non-negative. From the previous inequality applied to |f| we deduce

\sup_{s \ge 0}\, s\, \mu \left( \{ x \in X\, :\, |f(x) | \ge s \} \right)^{\frac{1}{q}}

\le 2^{1+q/2} C \mathcal{E}(|f|,|f|)^{1/2}

\le 2^{1+q/2} C \mathcal{E}(f,f)^{1/2}.

Theorem. (Sobolev inequality) Assume β > 1. There exists a constant C > 0 such that for every f ∈ ℱ,

\| f \|_{L^q(X,\mu)} \le C \mathcal{E}(f,f)^{1/2}

where q = 2β/(β – 1).

To show that the weak type inequality implies the desired Sobolev inequality, we will need another slicing argument and the following lemma is needed.

Lemma. For f ∈ ℱ, f ≥ 0, denote fk = (f – 2k)+ ∧ 2k, k ∈ ℤ. There exists a constant C > 0 such that for every f ∈ ℱ,

\sum_{k \in \mathbb{Z}} \mathcal{E} (f_k,f_k) \le C \mathcal{E} (f,f).

Proof: Let pt(x,y) denote the heat kernel of the semigroup Pt. We first observe that, once we prove

\sum_{k \in \mathbb{Z}} \int_X \int_X |f_{k} (x)-f_{k}(y)|^2 p_t(x,y) d\mu(x)d\mu(y) \le C\int_X \int_X |f (x)-f(y)|^2 p_t(x,y) d\mu(x)d\mu(y)

where C > 0 is independent from t, then

\liminf_{t \to 0^+} \sum_{k \in \mathbb{Z}} t^{-1 } \!\!\!\int_X \int_X |f_{\rho} (x)-f_{\rho}(y)|^2 p_t(x,y)d\mu(x)d\mu(y)

\le C\liminf_{t \to 0^+} t^{-1 } \!\!\!\int_X \int_X |f (x)-f(y)|^p p_t(x,y) d\mu(x)d\mu(y),

and, using the superadditivity of the liminf, one concludes

\sum_{k \in \mathbb{Z}} \liminf_{t \to 0^+} t^{-1 } \int_X \int_X |f_{k} (x)-f_{k}(y)|^2 p_t(x,y) d\mu(x)d\mu(y)

\le C \liminf_{t \to 0^+} t^{-1 } \int_X \int_X |f (x)-f(y)|^2 p_t(x,y) d\mu(x)d\mu(y)

which yields

\sum_{k \in \mathbb{Z}} \mathcal{E} (f_k,f_k) \le C \mathcal{E} (f,f).

We therefore aim to prove the inequality above. For each k ∈ ℤ, set Bk = {x ∈ X : 2k < f ≤ 2k+1}. In this way, the external integral on the left-hand side is decomposed it into an integral over Bk and Bkc. For the integrals over Bk, since the mapping f → fk is a contraction, it follows that

\sum_{k\in\mathbb{Z}}\int_{B_k} \int_X |f_{k} (x)-f_{k}(y)|^2 p_t(x,y)d\mu(x)d\mu(y)\leq \int_X \int_X |f (x)-f(y)|^2 p_t(x,y)d\mu(x)d\mu(y).

To perform the integrals over Bkc, we decompose them as

\sum_{k\in\mathbb{Z}}\int_{B_k^c} \int_{B_k} |f_{k} (x)-f_{k}(y)|^2 p_t(x,y) d\mu(x)d\mu(y) +\sum_{k\in\mathbb{Z}}\int_{B^c_k} \int_{B_k^c}|f_{k} (x)-f_{k}(y)|^2 p_t(x,y)d\mu(x)d\mu(y)

=:\sum_{k\in\mathbb{Z}}J_1(k)+\sum_{k\in\mathbb{Z}}J_2(k).

Again, the contraction property of f → fk yields

\sum_{k\in\mathbb{Z}}J_1(k)\leq \sum_{k\in\mathbb{Z}} \int_X \int_{B_k}|f_{k} (x)-f_{k}(y)|^2 p_t(x,y)d\mu(x)d\mu(y)

\leq \int_X \sum_{k\in\mathbb{Z}} \int_{B_k}|f_{k} (x)-f_{k}(y)|^2 p_t(x,y)d\mu(x)d\mu(y)\leq \int_X \int_X |f (x)-f(y)|^2 p_t(x,y) d\mu(x)d\mu(y).

On the other hand, notice that for any (x,y) ∈ Bkc × Bkc we have |fk(x) – fk(y)| ≠ 0 only if

(x,y)\in \{f(x)\leq 2^k<f(y)/2 \}\cup\{f(y)\leq 2^k<f(x)/2\}=:Z_k\cup Z_k^*.

Also, |fk(x) – fk(y)| = 2k for (x,y) ∈ Zk ∪ Zk*. Thus,

\sum_{k\in\mathbb{Z}}J_2(k)\leq \sum_{k\in \mathbb{Z}}\int_X\int_X\big(\mathbf{1}_{Z_k}(x,y)+\mathbf{1}_{Z^*_k}(x,y)\big)|f_k(x)-f_k(y)|^2 p_t(x,y)d\mu(x)d\mu(y)

=\int_X\int_X\sum_{k\in\mathbb{Z}}\big(\mathbf{1}_{Z_k}(x,y)+\mathbf{1}_{Z^*_k}(x,y)\big)2^{2k} p_t(x,y) d\mu(x)d\mu(y).

One can see that

\sum_{k\in\mathbb{Z}}\mathbf{1}_{Z_k}(x,y) 2^{2k}\leq 2 |f(x)-f(y)|^2

and the same holds for Zk*, hence

\sum\limits_{k\in\mathbb{Z}}J_1(k)+\sum\limits_{k\in\mathbb{Z}}J_2(k)\leq 5 \int_X\int_X|f(x)-f(y)|^2 p_t(x,y) d\mu(x)d\mu(y).

Adding to these the term above finally yields the result

We can now conclude the proof of the Sobolev inequality.

Proof of the Sobolev inequality: Let f ∈ ℱ. We can assume f ≥ 0. As before, denote fk = (f – 2k)+ ∧ 2k, k ∈ ℤ. From Lemma 13.2 applied to fk, we see that

\sup_{s \ge 0} s \mu \left( \{ x \in X\, :\, | f_k(x) | \ge s \} \right)^{\frac{1}{q}} \le C \mathcal{E}(f_k,f_k)^{1/2}.

In particular for s = 2k, we get

2^k \mu \left( \{ x \in X\, :\, f (x) \ge 2^{k+1} \} \right)^{\frac{1}{q}} \le C \mathcal{E}(f_k,f_k)^{1/2}

Therefore,

\sum_{k \in \mathbb{Z}} 2^{k q} \mu \left( \{ x \in X\, :\, f (x) \ge 2^{k+1} \} \right) \le C^q \sum_{k \in \mathbb{Z}} \mathcal{E}(f_k,f_k)^{q/2}.

Since q ≥ 2, one has

\sum_{k \in \mathbb{Z}} \mathcal{E}(f_k,f_k)^{q/2} \le \left( \sum_{k \in \mathbb{Z}} \mathcal{E}(f_k,f_k) \right)^{q/2}

Thus, from the previous lemma

\sum_{k \in \mathbb{Z}} 2^{k q} \mu \left( \{ x \in X\, :\, f (x) \ge 2^{k+1} \} \right)\le C \mathcal{E}(f,f)^{q/2}.

Finally, we observe that

\sum_{k \in \mathbb{Z}} 2^{k q} \mu \left( \{ x \in X\, :\, f (x) \ge 2^{k+1} \} \right) \ge \frac{q}{2^{q+1}-2^q} \sum_{k \in \mathbb{Z}}\int_{2^{k+1}}^{2^{k+2}} s^{q-1} \mu \left( \{ x \in X\, :\, f (x) \ge s \} \right)ds

\ge \frac{1}{2^{q+1}-2^q} \| f \|_{L^q(X,\mu)}^q.

The proof is thus complete.

Gagliardo-Nirenberg inequalities

Using similar methods (see the paper) in the general case β > 0 one can get  the family of Gagliardo-Nirenberg inequalities.

Theorem. Let q = 2β/(β – 1) with the convention that q = ∞ if β = 1. Let r,s ∈ (0,+∞] and θ ∈ (0,1] satisfying

\frac{1}{r}=\frac{\theta}{q}+\frac{1-\theta}{s}.

If β = 1, we assume r < +∞. Then, there exists a constant C > 0 such that for every f ∈ ℱ,

\| f \|_{L^r(X,\mu)} \le C \mathcal{E}(f,f)^{\theta/2}\|f\|_{L^s(X,\mu)}^{1-\theta}.

We explicitly point out some particular cases of interest.

  1. Assume that β > 1. If r = s, then r = 2β/(β – 1) and above recovers the Sobolev inequality

    \| f \|_{L^r(X,\mu)} \le C \mathcal{E}(f,f)^{1/2}.

  2. Assume that β > 1. If s = +∞ and r ≥ 2β/(β – 1), then above yields

    \| f \|_{L^r(X,\mu)} \le C \mathcal{E}(f,f)^{\theta/2} \| f \|^{1-\theta}_{L^\infty(X,\mu)}

    with θ = 2β/(r(β – 1)).

  3. If r = 2 and s = 1, then above yields the Nash inequality

    \| f \|_{L^2(X,\mu)} \le C \mathcal{E}(f,f)^{\theta/2} \| f \|^{1-\theta}_{L^1(X,\mu)}

    with θ = β/(1 + β).

In the case β = 1 one obtains the Trudinger-Moser inequalities.

Corollary . Assume that β = 1. Then, there exist constants c,C > 0 such that for every f ∈ ℱ with ℰ(f,f) = 1,

\int_X \left( \exp \left( c |f|^2 \right)-1 \right) d\mu \le C \| f \|^2_{L^2(X,\mu)}.

Posted in Dirichlet forms at NYU | Tagged , , | Leave a comment

Lecture 5. Examples of Dirichlet spaces

Contents

Riemannian manifolds

Let (M,g) be an n-dimensional Riemannian manifold with Riemannian volume measure μ and Riemannian distance d. We consider the quadratic form on M, which is obtained as the closure in L2(M,μ) of the quadratic form

M ⟨∇f, ∇g⟩ dμ, f,g ∈ Cc(M).

The domain of the closure is the Sobolev space W01,2(M) and its generator Δ is a self-adjoint extension of the Laplace-Beltrami operator. If the manifold is complete (which is equivalent to the metric space (M,d) being complete) then is the unique closed extension and W01,2(M) = W1,2(M). If we assume further that the Ricci curvature of M is bounded from below then the domain of Δ is the Sobolev space W2,2(M). If we even assume further that the Ricci curvature of M is non-negative, it is a well-known result by Li and Yau that the heat semigroup Pt admits a smooth heat kernel function pt(x,y) on [0,∞) × M × M for which there are constants c1, c2, C > 0 such that whenever t > 0 and x,y ∈ X,

\frac{C^{-1}}{\mu(B(x,\sqrt t))} \exp \left(-\frac{ c_1 d(x,y)^2}{t}\right)\le p_t(x,y)\le \frac{C}{\mu(B(x,\sqrt t))} \exp \left(-\frac{c_2d(x,y)^2}{(t}\right).

Carnot groups

A Carnot group of step N is a simply connected Lie group G whose Lie algebra can be stratified as follows:

g = 𝒱1 ⊕ … ⊕ 𝒱N,

where

[𝒱i , 𝒱j] = 𝒱i+j

and

𝒱s = 0, for s > N.

From the above properties, Carnot groups are nilpotent. The number

Q=\sum_{i=1}^N i \dim \mathcal{V}_{i}

is called the homogeneous dimension of G.

Let V1,…,Vd be a basis of the vector space 𝒱1. The vectors Vi‘s can be seen as left invariant vector fields on G. The left invariant sub-Laplacian on G is the operator:

L=\sum_{i=1}^d V_i^2

It is hypoelliptic and essentially self-adjoint on the space of smooth and compactly supported function f : G → ℝ with the respect to the Haar measure μ of G. The heat semigroup (Pt)t≥0 on G, defined through the spectral theorem, is then seen to be a Markov semigroup. By hypoellipticity of L, this heat semigroup admits a heat kernel denoted by pt(g,g’). It is then known that pt satisfies the double-sided Gaussian bounds:

\frac{C^{-1}}{t^{Q/2}} \exp \left(-\frac{ c_1 d(x,y)^2}{t}\right)\le p_t(x,y)\le \frac{C}{t^{Q/2}} \exp \left(-\frac{c_2d(x,y)^2}{(t}\right).

for some constants C,c1,c2 > 0. Here d(g,g’) denotes the Carnot-Carathéodory distance from g to g’ on G which is defined by

d(g,g')=\sup \left\{ |f(g)-f(g')| , \quad \sum_{i=1}^d (V_i f)^2 \le 1 \right\}.

Sierpiński gasket

A large class of examples for which Dirichlet form theory is useful is the class of p.c.f. fractals. For the sake of presentation we illustrate in detail the case of the Sierpiński gasket, which is one of the most popular examples of a p.c.f. fractal.

One of the classical ways to define the Sierpiński gasket is as follows:
let V0 = {p1, p2, p3} be a set of vertices of an equilateral triangle of side 1 in . Define

fi(z) = (z-pi)/2 + pi , for i = 1,2,3.

The Sierpiński gasket K is the unique non-empty compact subset in such that

K=f_1(K) \cup f_2(K) \cup f_3(K)

The set V0 is called the boundary of K, we will also denote it by ∂K. The Hausdorff dimension of K with respect to the Euclidean metric (denoted d(x,y) = |x – y| ) is given by dh = ln(3)/ln(2). A (normalized) Hausdorff measure on K is given by the Borel measure μ on K such that for every i1, …, in ∈ {1,2,3} ,

μ(fi1 ∘ … ∘ fin (K)) = 3-n.

This measure μ is dh-Ahlfors regular, i.e., there exist constants c,C > 0 such that for every x ∈ K and r ∈ [0, diam(K)],

crdh ≤ μ(B(x,r)) ≤ C rdh.

It will be useful to approximate the gasket K by a sequence of discrete objects. Namely, starting from the set V0 = {p1, p2, p3}, we define a sequence of sets {Vm}m≥0 inductively by

V_{m+1}=\bigcup_{i=1}^3 f_i(V_m)

Then we have a natural sequence of Sierpiński gasket graphs (or pre-gaskets) {Gm}m≥0 whose edges have length 2-m and whose set of vertices is Vm. Notice that #Vm = 3(3m + 1)/2. We will use the notations V* = ⋃m ≥ 0 Vm and V*0 = ⋃m≥0 Vm \ V0.

The Dirichlet form on the metric space K is defined by approximation. Let m ≥ 1. For any f ∈ ℝVm, we consider the quadratic form

\mathcal E_m(f,f)= \left(\frac53\right)^m\sum_{p,q\in V_m, p\sim q} (f(p)-f(q))^2

where p \sim q means that p,q are neighbors in the graph Vm.

We can then define a pre-Dirichlet form (ℰ, ℱ*) on V* by setting

* = {f ∈ ℝV* , limm → ∞m(f,f) < ∞}

and for f ∈ ℱ*

ℰ(f,f) = limm → ∞m(f,f).

Each function f ∈ ℱ* can be uniquely extended into a continuous function defined on K. We denote by the set of functions with such extensions. (ℰ, ℱ) is a Dirichlet form on L2(K,μ). The generator of the Dirichlet form (ℰ, ℱ), denoted by Δ, corresponds to the Laplacian with Neumann boundary condition.

In this example we have a continuous heat kernel pt(x,y) satisfying, for some c1, c2, c3, c4 ∈ (0,∞) and dH ≥ 1, dW ∈ [2,+∞),

c_{1}t^{-d_{H}/d_{W}}\exp\biggl(-c_{2}\Bigl(\frac{d(x,y)^{d_{W}}}{t}\Bigr)^{\frac{1}{d_{W}-1}}\biggr) \le p_{t}(x,y)\leq c_{3}t^{-d_{H}/d_{W}}\exp\biggl(-c_{4}\Bigl(\frac{d(x,y)^{d_{W}}}{t}\Bigr)^{\frac{1}{d_{W}-1}}\biggr)

for μ × μ-a.e. (x,y) ∈ X × X and each t ∈ (0,+∞). Here, dW = ln(5)/ln(3) is the so-called walk dimension of the Sierpiński gasket. A standard reference are the lecture notes by M. Barlow.

Cheeger metric measure spaces

Consider a locally compact, complete, metric measure space (X,d,μ) where μ is a Radon measure. Any open metric ball centered at x ∈ X with radius r > 0 will be denoted by

B(x,r) = {y ∈ X, d(x,y) < r}.

Definition. The measure μ is said to be doubling (VD) if there exists a constant C > 0 such that for every x ∈ X, r > 0,

0 < μ(B(x,2r)) ≤ C μ(B(x,r)) < +∞.

The Lipschitz constant of a function f ∈ Lip(X) is defined as

(\mathrm{Lip} f )(y):=\limsup_{r \to 0^+} \sup_{x \in X, d(x,y) \le r} \frac{|f(x)-f(y)|}{r}

Definition. The metric measure space (X,d,μ) is said to satisfy the 2-Poincaré inequality (P) if for any f ∈ Lip(X) and any ball B(x,R) of radius R > 0,

\int_{B(x,R)} | f(y) -f_{B(x,R)}|^2 d\mu (y) \le C R^2 \int_{B(x,\lambda R)} (\mathrm{Lip} f )(y)^2 d\mu (y)

where

f_{B(x,R)}:=\frac{1}{\mu(B(x,r))} \int_{B(x,R)} f(y) d\mu(y)
The constants C > 0 and λ ≥ 1 are independent from x, R and f.

Definition. A metric measure space satisfying (VD) and (P) is often called a Cheeger space (or PI space).

One can construct a “nice” Dirichlet form and Laplacian on any Cheeger space by the using the technique of Γ-convergence.

Definition. A sequence of forms {ℰn}n ≥ 1 is said to Mosco-converge to if

  1. For any sequence {fn}n≥1 ⊂ L2(X,μ) that converges *weakly* to f ∈ L2(X,μ) in L2(X,μ),

    liminfn→∞n(fn, fn) ≥ ℰ(f,f).

  2. For any f ∈ L2(X,μ) there exists a sequence {fn}n≥1 ⊂ L2(X,μ) that converges strongly to f in L2(X,μ) and

    limsupn→∞ ℰ(fn, fn) ≤ ℰ(f,f).

The idea is to consider Korevaar-Schoen type energy functionals defined for any f ∈ L2(X,μ) as

E(f,r):= \int_X\frac{1}{\mu(B(x,r))}\int_{B(x,r)} \frac{|f(y)-f(x)|^2}{r^{2}} d\mu(y) d\mu(x)

and the associated Korevaar-Schoen space

KS1,2(X) := {f ∈ L2(X,μ), limsupr → 0+ E(f,r) < +∞}.

One has then the following result:

Theorem. There exists a Dirichlet form (ℰ, ℱ) on L2(X,μ) such that:

  1. has domain ℱ = KS1,2(X);
  2. is a Γ-limit of E(f,rn), where rn is a positive sequence such that rn → 0;
  3. has a continuous heat kernel pt(x,y) that satisfies for t > 0 and x,y ∈ X,\frac{C^{-1}}{\mu(B(x,\sqrt t))} \exp \left(-\frac{ c_1 d(x,y)^2}{t}\right)\le p_t(x,y)\le \frac{C}{\mu(B(x,\sqrt t))} \exp \left(-\frac{c_2d(x,y)^2}{(t}\right).
Posted in Dirichlet forms at NYU | Tagged , , | Leave a comment

Lecture 4. The Lp theory of semigroups and diffusion operators as generators of Dirichlet forms

Contents

The Lp theory of heat semigroups

Our goal, in this section, is to define, for 1 ≤ p ≤ +∞, Pt on Lp := Lp(X,μ). This may be done in a natural way by using the Riesz-Thorin interpolation theorem that we recall below.

Theorem (Riesz-Thorin interpolation theorem)

Let 1 ≤ p0, p1, q0, q1 ≤ ∞, and θ ∈ (0,1). Define 1 ≤ p,q ≤ ∞ by

1/p = (1-θ)/p0 + θ/p1,     1/q = (1-θ)/q0 + θ/q1.

If T is a linear map such that

T : Lp0 → Lq0, ||T||Lp0→Lq0 = M0

T : Lp1 → Lq1, ||T||Lp1→Lq1 = M1,

then, for every f ∈ Lp0 ∩ Lp1,

||Tf||q ≤ M01-θM1θ ||f||p.

Hence T extends uniquely as a bounded map from Lp to Lq with

||T||Lp→Lq ≤ M01-θM1θ.

Remark The statement that T is a linear map such that

T : Lp0 → Lq0, ||T||Lp0→Lq0 = M0

T : Lp1 → Lq1, ||T||Lp1→Lq1 = M1

means that there exists a map T : Lp0 ∩ Lp1 → Lq0 ∩ Lq1 with

supf ∈ Lp0 ∩ Lp1 , ||f||p0 ≤ 1 ||Tf||q0 = M0

and

supf ∈ Lp0 ∩ Lp1 , ||f||p1 ≤ 1 ||Tf||q1 = M1.

In such a case, T can be uniquely extended to bounded linear maps T0 : Lp0 → Lq0 , T1 : Lp1 → Lq1. With a slight abuse of notation, these two maps are both denoted by T in the theorem.

Remark If f ∈ Lp0 ∩ Lp1 and p is defined by

1/p = (1-θ)/p0 + θ/p1,

then by Hölder’s inequality, f ∈ Lp and

||f||p ≤ ||f||p01-θ ||f||p1θ.

We now are in position to state the following theorem:

Theorem Let (Pt)t≥0 be a strongly continuous self-adjoint contraction Markovian semigroup on L2(X,μ). The space L1 ∩ L is invariant under Pt and Pt may be extended from L1 ∩ L to a contraction semigroup (Pt(p))t≥0 on Lp for all 1 ≤ p ≤ ∞: For f ∈ Lp,

||Ptf||Lp ≤ ||f||Lp.

These semigroups are consistent in the sense that for f ∈ Lp ∩ Lq,

Pt(p)f = Pt(q)f.

Proof

If f,g ∈ L1 ∩ L which is a subset of L1 ∩ L, then,

|∫X (Ptf)g dμ| = |∫X f (Ptg) dμ| ≤ ||f||L1 ||Ptg||L ≤ ||f||L1 ||g||L.

This implies

||Ptf||L1 ≤ ||f||L1.

The conclusion follows then from the Riesz-Thorin interpolation theorem.

Exercise Show that if f ∈ Lp and g ∈ Lq with 1/p + 1/q = 1 then,

n f Pt(q) g dμ = ∫n g Pt(p) f dμ.

Exercise 

  1. Show that for each f ∈ L1, the L1-valued map t → Pt(1)f is continuous.
  2. Show that for each f ∈ Lp, 1 < p < 2, the Lp-valued map t → Pt(p)f is continuous.
  3. Finally, by using the reflexivity of Lp, show that for each f ∈ Lp and every p ≥ 1, the Lp-valued map t → Pt(p)f is continuous.

We mention, that in general, the L valued map t → Pt(∞)f is not continuous.

Diffusion operators as generators of Dirichlet forms

Consider a diffusion operator

L = ∑i,j=1n σij(x) ∂2/∂xi∂xj + ∑i=1n bi(x) ∂/∂xi,

where bi and σij are continuous functions on n and for every x ∈ ℝn, the matrix ij(x))1 ≤ i,j ≤ n is a symmetric and non-negative matrix.

Assume that there is Borel measure μ on ℝn which is equivalent to the Lebesgue measure and that symmetrizes L in the sense that for every smooth and compactly supported functions f, g : ℝn → ℝ,

n gLf dμ = ∫n fLg dμ.

For instance, if one can write

Lf = -div(a ∇f),

where a is a smooth field of positive and symmetric matrices, then the Lebesgue measure symetrizes L. From a previous lemma  the quadratic form

ℰ(f,g) = -∫n g Lf dμ, f,g ∈ Cc(ℝn)

is closable. Let ℰ̄ denotes its closure in L2(ℝn,μ).

Proposition The quadratic form ℰ̄ is a Dirichlet form.

Proof

We need to prove that ℰ̄ is Markovian. It is enough to prove that if u ∈ ℱ = 𝒟(ℰ), then |u| ∈ ℱ with ℰ̄(|u|, |u|) ≤ ℰ̄(u,u) and that if u ∈ ℱ with u ≥ 0, then u ∧ 1 ∈ ℱ with ℰ̄(|u|,|u|) ≤ ℰ̄(u,u). We prove the first requirement, the second being established in a similar manner is let as an exercise to the reader.

Let u ∈ Cc(ℝn) and consider

ϕε(x) = (x2 + ε2)^{1/2}, ε > 0.

One can check that ϕε(u) → |u| in L2(ℝn,μ) and that ϕε(u) is a Cauchy sequence for the norm

||f||2 = ||f||2L2(ℝn,μ) + ℰ̄(f,f).

Since ℰ̄ is closed this implies that |u| ∈ ℱ and that ϕε(u) → |u| converges to u in the above norm.

Now, using chain rule we see that for every smooth function u ∈ Cc(ℝn),

ε(u)(x) ≥ u(x)/(u(x)2 + ε2)^{1/2} Lu(x).

Multiplying by ϕε(u) and integrating we get

ℰ(ϕε(u), ϕε(u)) ≤ ℰ(u,u)

Taking the limit ε → 0 yields

ℰ̄(|u|, |u|) ≤ ℰ̄(u,u)

The above inequality extends then to all u ∈ ℱ by using the density of Cc(ℝn) in the || · || norm and the closedness of ℰ.

Posted in Dirichlet forms at NYU | Tagged , | Leave a comment

Lecture 3. Markov semigroups and Dirichlet forms

Let (X, ℬ) be a measurable space. We say that (X, ℬ) is a good measurable space if there is a countable family generating and if every finite measure γ on (X × X, ℬ ⊗ ℬ) can be decomposed as

γ(dx dy) = k(x,dy) γ1(dx)

where γ1 is the projection of γ on the first coordinate and k is a kernel, i.e. k(x,·) is a finite measure on (X, ℬ) and x → k(x,A) is measurable for every A ∈ ℬ.

For instance, if X is a Polish space (or a Radon space) equipped with its Borel σ-field, then it is a good measurable space.

Throughout the lecture, we will consider (X, ℬ, μ) to be a good measurable space equipped with a σ-finite measure μ.

Contents

Markovian semigroups

Definition Let (Pt)t≥0 be a strongly continuous self-adjoint contraction semigroup on L2(X,μ). The semigroup (Pt)t≥0 is called Markovian if and only if for every f ∈ L2(X,μ) and t ≥ 0:

  1. f ≥ 0, a.e. ⇒ Ptf ≥ 0, a.e.

  2. f ≤ 1, a.e. ⇒ Ptf ≤ 1, a.e.

We note that if (Pt)t≥0 is Markovian, then for every f ∈ L2(X,μ) ∩ L(X,μ),

||Ptf||L(X,μ) ≤ ||f||L(X,μ).

As a consequence (Pt)t≥0 can be extended to a contraction semigroup defined on all of L(X,μ).

Definition A transition function {pt,t ≥ 0} on X is a family of kernelsnpt : X × ℬ → [0,1]

such that:

  1. For t ≥ 0 and x ∈ X, pt(x, ·) is a finite measure on X;
  2. For t ≥ 0 and A ∈ ℬ the application x → pt(x,A) is measurable;
  3. For s,t ≥ 0, a.e. x ∈ X and A ∈ ℬ,

    pt+s(x,A) = ∫X pt(y,A) ps(x,dy).

The relation above is often called the Chapman-Kolmogorov relation.

Theorem (Heat kernel measure)nLet (Pt)t≥0 be a strongly continuous self-adjoint contraction Markovian semigroup on L2(X,μ).nThere exists a transition function {pt,t ≥ 0} on X such that for every f ∈ L(X,μ) and a.e. x ∈ X

Ptf(x) = ∫X f(y) pt(x,dy) , t > 0.

This transition function is called the heat kernel measure associated to (Pt)t≥0.

The proof relies on the following lemma sometimes called the bi-measure theorem. A set function ν : ℬ ⊗ ℬ → [0,+∞) is called a bi-measure, if for every A ∈ ℬ, ν(A, ·) and ν(·, A) are measures.

Lemma If ν : ℬ ⊗ ℬ → [0,+∞) is a bi-measure, then there exists a measure γ on ℬ ⊗ ℬ such that for every A,B ∈ ℬ,

γ(A × B) = ν(A,B).

Proof of Theorem 7.3

We assume that μ is finite and let as an exercise the extension to σ-finite measures. For t > 0, we consider the set function

νt(A,B) = ∫X 1A Pt 1B dμ.

Since Pt is supposed to be Markovian, it is a bi-measure. From the bi-measure theorem, there exists a measure γt on ℬ ⊗ ℬ such that for every A,B ∈ ℬ,

γt(A × B) = νt(A,B) = ∫X 1A Pt 1B dμ.

The projection of γt on the first coordinate is (Pt1) dμ, thus from the measure decomposition theorem, γt can be decomposed as

γt(dx dy) = pt(x,dy) μ(dx)

for some kernel pt. One has then for every A,B ∈ ℬ

X 1A Pt 1B dμ = ∫AB pt(x,dy) μ(dx),

from which it follows that for every f ∈ L(X,μ), and a.e. x ∈ X

Ptf(x) = ∫X f(y) pt(x,dy).

The relation

pt+s(x,A) = ∫X pt(y,A) ps(x,dy)

follows from the semigroup property.

Exercise Prove Theorem 7.3 if μ is σ-finite.

Exercise Show that for every non-negative measurable function F : X × X → ℝ,

XX F(x,y) pt(x,dy) dμ(x) = ∫XX F(x,y) pt(y,dx) dμ(y).

Definition Let (Pt)t≥0 be a strongly continuous self-adjoint contraction Markovian semigroup on L2(X,μ). We say that the semigroup {Pt}t∈[0,∞) admits a heat kernel if the heat kernel measures have a density with respect to μ, i.e. there exists a measurable function p : ℝ>0 × X × X → ℝ≥0, such that for every t > 0, a.e. x,y ∈ X, f ∈ L(X,μ),

Ptf(x) = ∫X pt(x,y) f(y) dμ(y).

If the heat kernel exists, we will often denote p(t,x,y) as pt(x,y) for t > 0 and a.e. x,y ∈ X.

Dirichlet forms

Definition A function v on X is called a normal contraction of the function u if for almost every x, y ∈ X,

|v(x)-v(y)| ≤ |u(x) – u(y)| and |v(x)| ≤ |u(x)|.

Definition Let (ℰ,ℱ = dom(ℰ)) be a densely defined closed quadratic form on L2(X,μ). The form is called a Dirichlet form if it is Markovian, that is, has the property that if u ∈ ℱ and v is a normal contraction of u then v ∈ ℱ and

ℰ(v,v) ≤ ℰ(u,u).

Exercise Show that a densely defined closed quadratic form on L2(X,μ) is Markovian if and only if for every u ∈ ℱ, (0 ∨ u) ∧ 1 ∈ ℱ and ℰ( (0 ∨ u) ∧ 1, (0 ∨ u) ∧ 1 ) ≤ ℰ(u,u).

Theorem Let (Pt)t≥0 be a strongly continuous self-adjoint contraction semigroup on L2(X,μ). Then, (Pt)t≥0 is a Markovian semigroup if and only if the associated closed symmetric form on L2(X,μ) is a Dirichlet form.

Proof

Let (Pt)t≥0 be a strongly continuous self-adjoint contraction Markovian semigroup on L2(X,μ). There exists a transition function {pt, t ≥ 0} on X such that for every u ∈ L(X,μ) and a.e. x ∈ X

Ptu(x) = ∫X u(y) pt(x,dy), t > 0.

Denote

kt(x) = Pt1(x) = ∫X pt(x,dy).

We observe that from the Markovian property of Pt, we have 0 ≤ kt ≤ 1 a.e.
We have then

1/2 ∫XX (u(x) – u(y))2 pt(x,dy) dμ(x) = ∫X u(x)2 kt(x)dμ(x) – ∫X u(x) Ptu(x) dμ(x).

Therefore,

⟨u – Ptu, u⟩ = 1/2 ∫XX (u(x) – u(y))2 pt(x,dy) dμ(x) + ∫X u(x)2 (1 – kt(x)) dμ(x).

Let us now assume that u ∈ ℱ and that v is a normal contraction of u. One has

XX (v(x) – v(y))2 pt(x,dy) dμ(x) ≤ ∫XX (u(x) – u(y))2 pt(x,dy) dμ(x)

and

X v(x)2 (1 – kt(x)) dμ(x) ≤ ∫X u(x)2 (1 – kt(x)) dμ(x).

Therefore,

⟨v – Ptv,v⟩ ≤ ⟨u – Ptu,u⟩

Since u ∈ ℱ, one knows that (1/t)⟨u – Ptu,u⟩ converges to ℰ(u) when t → 0. Since (1/t)⟨v – Ptv,v⟩ is non-increasing and bounded it does converge when t → 0. Thus v ∈ ℱ and

ℰ(v) ≤ ℰ(u).

One concludes that is Markovian.

Now, consider a Dirichlet form ℰ and denote by Pt the associated semigroup in L2(X,μ) and by A its generator.
The main idea is to first prove that for λ > 0, the resolvent operator (λId – A)-1 preserves the positivity of function. Then, we may conclude by the fact that for f ∈ L2(X,μ), in the L2(X,μ) sense

Ptf = limn → +∞ (Id – t/n A)-nf.

Let λ > 0. We consider on the norm

||f||2λ = ||f||2L2(X,μ) + λℰ(f,f).

From the Markovian property of ℰ, if u ∈ ℱ, then |u| ∈ ℱ and

ℰ(|u|, |u|) ≤ ℰ(u, u).

We consider the bounded operator

Rλ = (Id – λA)-1

that goes from L2(X,μ) to 𝒟(A) ⊂ ℱ. For f ∈ ℱ and g ∈ L2(X,μ) with g ≥ 0, we have

⟨|f| , Rλ g⟩λ = ⟨|f| , Rλg⟩L2(X,μ) – λ⟨|f| , ARλ g⟩L2(X,μ)

= ⟨|f|, (Id – λA) Rλ g⟩L2(X,μ)

=⟨|f|, g⟩L2(X,μ)

≥ |⟨f, g⟩L2(X,μ)|

≥ |⟨f , Rλg⟩λ|.

Moreover, from inequality above, for f ∈ ℱ,

|||f|||λ2 = |||f| ||2L2(X,μ) + λℰ(|f|,|f|)

≤ ||f||2L2(X,μ) + λℰ(f,f)

≤ ||f||λ2.

By taking f = Rλ g in the two above sets of inequalities, we draw the conclusion

|⟨Rλg, Rλg⟩λ| ≤ ⟨|Rλg| , Rλg⟩λ ≤ |||Rλg|||λ ||Rλg||λ ≤ |⟨Rλg, Rλg⟩λ|.

The above inequalities are therefore equalities which implies

Rλg = |Rλg|.

As a conclusion if g ∈ L2(X,μ) is a.e. ≥ 0, then for every λ > 0, (Id – λA)-1g ≥ 0 a.e.. Thanks to the spectral theorem, in L2(X,μ),

Ptg = limn → +∞ (Id – t/n A)-ng.

By passing to a subsequence that converges pointwise almost surely, we deduce that Ptg ≥ 0 almost surely.

The proof of

f ≤ 1, a.e. ⇒ Ptf ≤ 1, a.e.

follows the same lines:

  1. The first step is to observe that if 0 ≤ f ∈ ℱ, then 1 ∧ f ∈ ℱ and moreover

ℰ(1 ∧ f, 1 ∧ f) ≤ ℰ(f,f).

2. Let f ∈ L2(X,μ) satisfy 0 ≤ f ≤ 1 and set g = Rλf = (Id – λA)-1f ∈ ℱ and h = 1 ∧ g. According to the first step, h ∈ ℱ and ℰ(h,h) ≤ ℰ(g,g). Now, we observe that:

||g – h||λ2 = ||g||λ2 – 2⟨g,h⟩λ + ||h||λ2

= ⟨Rλf,f⟩L2(X,μ) – 2⟨f,h⟩L2(X,μ) + ||h||2L2(X,μ) + λℰ(h,h)

= ⟨Rλf,f⟩L2(X,μ) – ||f||2L2(X,μ) + ||f – h||2L2(X,μ) + λℰ(h,h)

≤ ⟨Rλf,f⟩L2(X,μ) – ||f||2L2(X,μ) + ||f – g||2L2(X,μ) + λℰ(g,g) = 0.

As a consequence g = h, that is 0 ≤ g ≤ 1.

3. The previous step shows that if f ∈ L2(X,μ) satisfies 0 ≤ f ≤ 1 then for every λ > 0, 0 ≤ (Id – λL)-1 f ≤ 1. Thanks to the spectral theorem, in L2(X,μ),

Ptf = limn → +∞ (Id – t/n L)-nf.

By passing to a subsequence that converges pointwise almost surely, we deduce that 0 ≤ Ptf ≤ 1 almost surely.

Posted in Dirichlet forms at NYU | Tagged , | Leave a comment

Lecture 2. Quadratic forms in Hilbert spaces

Contents

Quadratic forms and generators

Definition A quadratic form on H is a non-negative definite, symmetric bilinear form 𝒟(ℰ) × 𝒟(ℰ) → ℝ, where 𝒟(ℰ) is a dense subspace of H. A quadratic form on H is said to be closed if 𝒟(ℰ) equipped with the norm

‖f‖𝒟(ℰ)2= ‖f‖2 + ℰ(f,f)

is a Hilbert space. A quadratic form on H is said to be closable it admits a closed extension, i.e. there exists a closed quadratic form ℰ’ such that 𝒟(ℰ) ⊂ 𝒟(ℰ’) and ℰ’ coincides with on 𝒟(ℰ) × 𝒟(ℰ).

Lemma A quadratic form ℰ is closable if and only if for any sequence fn in 𝒟(ℰ) such that fn → 0 in H and ℰ(fn – fm, fn – fm) → 0 when n, m → +∞ one has ℰ(fn, fn) → 0.

Proof

On 𝒟(ℰ), let us consider the following norm

||f||2 = ||f||2 + ℰ(f, f).

By completing 𝒟(ℰ) with respect to this norm, we get an abstract Hilbert space (H,⟨⋅,⋅⟩). Since for f ∈ 𝒟(ℰ), ||f|| ≤ ||f||, the injection map ι : (𝒟(ℰ), || · ||) → (H, || · ||) is continuous and it may therefore be extended into a continuous map ῑ : (H, || · ||) → (H, || · ||). Let us show that ̄ι is injective so that H may be identified with a subspace of H. So, let f ∈ H such that ῑ(f) = 0. We can find a sequence fn ∈ 𝒟(ℰ), such that || fn – f || → 0 and || fn || → 0. We have then

||f||2 = limn → + ∞ ⟨fn, fn

= limn → + ∞ ⟨fn,fn⟩ + ℰ(fn, fn)

= 0,

thus f=0 and is injective. Therefore, H may be identified with a subspace of H and the quadratic form on H defined by

ℰ'(f) = ||f||2 – ||f||2, f ∈ H

is closed because (H,⟨⋅,⋅⟩) is a Hilbert space and obviously is an extension of .

If a quadratic form is closable, then its minimal closed extension is called the closure of . In that case, one can easily check that the closure of is actually the quadratic form constructed in the previous proof.

Theorem Let be a closed symmetric non-negative bilinear form on H. There exists a unique densely defined non-positive self-adjoint operator A on H defined by

𝒟(A) = {f ∈ H, ∃g ∈ H, ∀h ∈ H, ℰ(f,h) = -⟨h,g⟩}

Af = g.

The operator A is called the generator of ℰ. Conversely, if A is a densely defined non-positive self-adjoint operator on H, one can define a closed symmetric non-negative bilinear form ℰ on H by

𝒟(ℰ) = 𝒟((-A)1/2), ℰ(f,g) = ⟨(-A)1/2f,(-A)1/2g⟩.

Proof

Let ℰ be a closed symmetric non-negative bilinear form on H. As usual, we denote by ℱ the domain of ℰ. We note that for λ > 0, ℱ equipped with the norm (||f||2 + λℰ(f))1/2 is a Hilbert space because ℰ is closed. From the Riesz representation theorem, there exists then a linear operator Rλ :H → ℱ such that for every f ∈ H, g ∈ ℱ

⟨f,g⟩ = λ⟨Rλ f , g⟩ + ℰ(Rλ f,g).

From the definition, the following properties are then easily checked:

  1. ||Rλf|| ≤ (1/λ) ||f|| (apply the definition of Rλ with g = Rλf and then use the Cauchy-Schwarz inequality);
  2. For every f,g ∈ H, ⟨Rλf , g⟩ = ⟨f , Rλg⟩;
  3. Rλ1 – Rλ2 + (λ1 – λ2)Rλ1Rλ2 = 0;
  4. For every f ∈ H, limλ → +∞ || λRλf -f || = 0.

We then claim that Rλ is invertible. Indeed, if Rλf = 0, then for α > λ , one has from 3, Rαf = 0. Therefore f = limα → +∞ Rαf = 0. Denote then

Af = λf – Rλ-1f,

and 𝒟(A) is the range of Rλ. It is straightforward to check that A does not depend on λ. The operator A is a densely defined self-adjoint operator that satisfies the properties stated in the theorem (Exercise !).

Conversely, if A is a densely defined non-positive self-adjoint operator on H, then (-A)1/2 is a densely defined self-adjoint operator and the quadratic form

ℰ(f,g) := ⟨(-A)1/2f, (-A)1/2g⟩

is closed and densely defined on 𝒟((-A)1/2).

Exercise Prove the properties 1,2,3,4 of the previous proof.

In practice, the following lemma is often useful to construct closed quadratic forms and easily follows from the previous results.

Lemma Let A be a densely defined non-positive symmetric operator 𝒟(A) → H. The quadratic form

ℰ(f,g) = -⟨f, Ag⟩, f,g ∈ 𝒟(A)

is closable and the generator of its closure is a self-adjoint extension of A.

Semigroups and quadratic forms

Theorem Let (Pt)t≥0 be a strongly continuous self-adjoint contraction semigroup on H. One can define a closed quadratic form on H by

ℰ(f,f) := limt → 0 ⟨(Id – Pt)/t f, f⟩,

where the domain of this form is the set of f‘s for which the limit exists. The quadratic form is called the quadratic form associated to the semigroup (Pt)t≥0.

Proof

Let A be the generator of the semigroup (Pt)t ≥ 0. We use spectral theorem to represent A as

U-1 A U g(x) = -λ(x) g(x),

so that

U-1 Pt U g(x) = e-tλ(x) g(x).

We then note that for every g ∈ L2(Ω,ν),

⟨(Id – Pt)/t Ug, Ug⟩ = ∫Ω (1 – e-tλ(x))/t g(x)2 dν(x).

This proves that for every f ∈ H, the map t → ⟨(Id – Pt)/t f, f⟩ is non-increasing. Therefore, the limit limt → 0 ⟨(Id – Pt)/t f, f⟩ exists if and only if ∫Ω (U-1f)2(x) λ(x) dν(x) < +∞, which is equivalent to the fact that f ∈ 𝒟((-A)1/2). In which case we have

limt → 0 ⟨(Id – Pt)/t f, f⟩ = ||(-A)1/2f||2.

Since (-A)1/2 is a densely defined self-adjoint operator, the quadratic form

ℰ(f) := ||(-A)1/2f||2

is closed and densely defined on ℱ := 𝒟((-A)1/2).

A first example: The Dirichlet energy on an open set Ω ⊂ ℝn

Let Ω ⊂ ℝn be an open connected set. We do not assume any regularity on the boundary of Ω. Classically, one can define the (1,2) Sobolev space

W1,2(Ω) = {f ∈ L2(Ω) : ∂f/∂xi ∈ L2(Ω)}

where the derivatives ∂u/∂xi are understood in the weak sense. The quadratic form

ℰ(f,g) = ∫Ω ⟨∇f, ∇g⟩ dx = ∑i=1nΩ ∂f/∂xi ∂g/∂xi dx

with domain W1,2(Ω) is then a closed densely defined quadratic form on L2(Ω) since it is well-known that the Sobolev norm

||f||2W1,2(Ω) = ||f||2L2(Ω) + ||∇f||2L2(Ω)

is complete. The generator of the form is called the Neumann Laplacian on Ω.

On the other hand, let

Δ = ∑i=1n2/∂xi2

be the usual Laplacian on n, the derivatives being understood in the ordinary sense, and Cc(Ω) be the set of smooth functions with a compact support included in Ω. Then, from a lemma, the quadratic form

0(f,g) = -∫Ω f Δg dx

with domain Cc(Ω) is closable. The domain of the closure of 0 is the Sobolev space W01,2(Ω) and the generator of the closure of 0 is called the Dirichlet Laplacian on Ω.

Notice that both the Neumann and the Dirichlet Laplacian are self-adjoint extensions of the Laplacian Δ with domain Cc(Ω). In general, the Neumann and Dirichlet Laplacian do not coincide. For instance if the boundary of Ω is smooth, then smooth functions in the domain of the Neumann Laplacian have vanishing normal derivatives while smooth functions in the domain of the Dirichlet Laplacian vanish on the boundary of u.

Posted in Dirichlet forms at NYU | Tagged , | Leave a comment