L3

The interplay between stochastic processes and optimal control has been extensively explored in the literature. With the recent surge in the use of diffusion models, stochastic processes have increasingly been applied to sample generation. This talk builds on the log transform, known as the Cole-Hopf transform in Brownian motion contexts, and extends it within a more abstract framework that includes a linear operator. Within this framework, we found that the well-known relationship between the Cole-Hopf transform and optimal transport is a particular instance where the linear operator acts as the infinitesimal generator of a stochastic process. We also introduce a novel scenario where the linear operator is the adjoint of the generator, linking to Bayesian inference under specific initial and terminal conditions. Leveraging this theoretical foundation, we develop a new algorithm, named the HJ-sampler, for Bayesian inference for the inverse problem of a stochastic differential equation with given terminal observations. The HJ-sampler involves two stages: solving viscous Hamilton-Jacobi (HJ) partial differential equations (PDEs) and sampling from the associated stochastic optimal control problem. Our proposed algorithm naturally allows for flexibility in selecting the numerical solver for viscous HJ PDEs. We introduce two variants of the solver: the Riccati-HJ-sampler, based on the Riccati method, and the SGM-HJ-sampler, which utilizes diffusion models. Numerical examples demonstrate the effectiveness of our proposed methods. This is an ongoing joint work with Zongren Zou, Jerome Darbon, and George Em Karniadakis.

We present a spectral method that converges exponentially for a variety of fractional integral equations on a closed interval. The method uses an orthogonal fractional polynomial basis that is obtained from an appropriate change of variable in classical Jacobi polynomials. For a problem arising from time-fractional heat and wave equations, we elaborate the complexities of three spectral methods, among which our method is the most performant due to its superior stability. We present algorithms for building the fractional integral operators, which are applied to the orthogonal fractional polynomial basis as matrices. 

The main component of flow machines is the rotor; however, there may also be stationary parts surrounding the rotor, which are the diffuser blades. In order to consider these two parts simultaneously, the most intuitive approach is to perform a transient flow simulation; however, the computational cost is relatively high. Therefore, one possible approach is the Multiple Reference Frame (MRF) approach, which considers two directly coupled zones: one for the rotating reference frame (for the rotor blades) and one for the stationary reference frame (for the diffuser blades). When taking into account topology optimisation, some changes are required in order to take both rotating and stationary parts simultaneously in the design, which also leads to changes in the composition of the multi-objective function. Therefore, the topology optimisation method is formulated for MRF while also proposing this new multi-objective function. An integer variable-based optimisation algorithm is considered, with some adjustments for the MRF case. Some numerical examples are presented.

Our research explores the application of reinforcement learning (RL) strategies to solve complex combinatorial research problems, specifically the Job-shop Scheduling Problem (JSP) and the Stochastic Vehicle Routing Problem with Time Windows (SVRP). For JSP, we utilize Curriculum Learning (CL) to enhance the performance of dispatching policies. This approach addresses the significant optimality gap in existing end-to-end solutions by structuring the training process into a sequence of increasingly complex tasks, thus facilitating the handling of larger, more intricate instances. Our study introduces a size-agnostic model and a novel strategy, the Reinforced Adaptive Staircase Curriculum Learning (RASCL), which dynamically adjusts difficulty levels during training, focusing on the most challenging instances. Experimental results on Taillard and Demirkol datasets show that our approach reduces the average optimality gap to 10.46% and 18.85%, respectively.

For SVRP, we propose an end-to-end framework employing an attention-based neural network trained through RL to minimize routing costs while addressing uncertain travel costs and demands, alongside specific customer delivery time windows. This model outperforms the state-of-the-art Ant-Colony Optimization algorithm by achieving a 1.73% reduction in travel costs and demonstrates robustness across diverse environmental settings, making it a valuable baseline for future research. Both studies mark advancements in the application of machine learning techniques to operational research.

The Fokker-Planck equation is one of the major tools of statistical physics in the description of stochastic processes, with numerous applications in physics, chemistry and biology. In the case of heterogeneous diffusions, the formulation of the equation depends on the choice of the discretization of the stochastic integral in the underlying Langevin-equation due to the multiplicative noise. In the Fokker-Planck equation, the choice of discretization then enters as a parameter in the definition of drift and diffusion terms. I show how both short- and long-time limits are affected by this choice. In the long-time limit, the existence of normalizable probability distribution functions is not always guaranteed which can be remedied by invoking elements of infinite ergodic theory. 

[1] S. Giordano, F. Cleri, R. Blossey, Phys Rev E 107, 044111 (2023)

[2] T. Dupont, S. Giordano, F. Cleri, R. Blossey, arXiv:2401.01765 (2024)

This will be an informal discussion seminar based on https://arxiv.org/abs/2211.07592:

The constrained Hamiltonian analysis of geometric actions is worked out before applying the construction to the extended Bondi-Metzner-Sachs group in four dimensions. For any Hamiltonian associated with an extended BMS4 generator, this action provides a field theory in two plus one spacetime dimensions whose Poisson bracket algebra of Noether charges realizes the extended BMS4 Lie algebra. The Poisson structure of the model includes the classical version of the operator product expansions that have appeared in the context of celestial holography. Furthermore, the model reproduces the evolution equations of non-radiative asymptotically flat spacetimes at null infinity.

Positive logic is a generalisation of full first-order logic, where negation is not built in, but can be added as desired. In joint work with Jan Dobrowolski we succesfully generalised the recent development on Kim-independence in NSOP1 theories to the positive setting. One of the important theorems in this development is the independence theorem, whose statement is very similar to the well-known statement for simple theories, and allows us to amalgamate independent types. In this talk we will have a closer look at the proof of this theorem, and what needs to be changed to make the proof work in positive logic compared to full first-order logic.

Causal optimal transport and the related adapted Wasserstein distance have recently been popularized as a more appropriate alternative to the classical Wasserstein distance in the context of stochastic analysis and mathematical finance. In this talk, we establish some interesting consequences of causality for transports on the space of continuous functions between the laws of stochastic differential equations.
 

We first characterize bicausal transport plans and maps between the laws of stochastic differential equations. As an application, we are able to provide necessary and sufficient conditions for bicausal transport plans to be induced by bi-causal maps. Analogous to the classical case, we show that bicausal Monge transports are dense in the set of bicausal couplings between laws of SDEs with unique strong solutions and regular coefficients.

 This is a joint work with Rama Cont.