Laboratoire de Finance des Marchés de l'Energie

Projet de recherche commun entre l’Université Paris Dauphine, l'école Polytechnique, le Centre de Recherche en Economie et en Statistique (CREST) et la R&D du groupe EDF.

Créé en même temps que la Chaire Dauphine Ecole Polytechnique EDF - CA CIB« Finance et Développement Durable – Approches Quantitatives », il a pour vocation d’accueillir des chercheurs de toutes institutions académiques désireux de travailler avec des ingénieurs-chercheurs de la R&D d'EDF sur les problématiques d’économie mathématique et de finance quantitative de long terme du secteur énergétique.


Conférence ANR CAESARS 5-7 September 2018 

In electrical system, strong evolutions are under way that will change deeply the organisation of the whole sector in the short and long term horizon: quick development of renewable technology, volatile and unpredictable production, costly investments in a difficult economic context, competing environment, emission market, new usages of electric vehicles... The simulation and analysis of the evolution of the electrical system is fundamental for better sustaining the energetic transition to a Clean Energy World but it is challenging because of various sources of uncertainties and their interdependence, the interaction between actors and between decisions. The aim of this event is to gather practitioners and academic people, experts in mathematical modeling and numerical simulation applied to energy, to present the most recent advances in these fields.


Open-source stochastic optimization library

The STochastic OPTimization library (StOpt) aims at providing tools in C++ for solving some stochastic optimization problems encountered in finance or in the industry. A python binding is available for some C++ objects provided permitting to easily solve an optimization problem by regression.

Different methods are available : dynamic programming methods based on Monte Carlo  with regressions (global, local and  sparse regressors), for underlying states following;  some uncontrolled Stochastic Differential Equations  ;  Semi-Lagrangian methods for Hamilton Jacobi Bellman general equations for underlying states following some controlled  Stochastic Differential Equations  and Stochastic Dual Dynamic Programming methods to deal with stochastic stocks management problems in high dimension

Derniers rapports de recherche


Some machine learning schemes for high-dimensional nonlinear PDEs - C. HURE, H. PHAM, X. WARIN

We propose new machine learning schemes for solving high dimensional nonlinear partial differential equations (PDEs). Relying on the classical backward stochastic differential equation (BSDE) representation of PDEs, our algorithms estimate simultaneously the solution and its gradient by deep neural networks. These approximations are performed at each time step from the minimization of loss functions...

Deep neural networks algorithms for stochastic control problems on nite horizon, Part 2: numerical applications - A. Bachouch, C. Huré, N. Langrené, H. Pham

This paper presents several numerical applications of deep learning-based algorithms that have been analyzed in . Numerical and comparative tests using TensorFlow illustrate the performance of our di erent algorithms, namely control learning by performance iteration (algorithms NNcontPI and ClassifPI), control learning by hybrid iteration (algorithms Hybrid-Now and Hybrid-LaterQ), on the 100-dimensional nonlinear PDEs examples...

Deep neural networks algorithms for stochastic control problems on finite horizon, part I : convergence analysis - C. Hure, H. Pham, A. Bachouch and N. Langrené

This paper develops algorithms for high-dimensional stochastic control problems based on deep learning and dynamic programming (DP). Di erently from the classical approximate DP approach, we rst approximate the optimal policy by means of neural networks in the spirit of deep reinforcement learning, and then the value function by Monte Carlo regression. This is achieved...

Prochains séminaires - 14h à l'IHP