Project description
Nonlinear Monte Carlo type methods for high-dimensional approximation problems
In many relevant real-world problems it is of fundamental importance to approximately compute evaluations of high-dimensional functions. Standard deterministic approximation methods often suffer in this context from the so-called curse of dimensionality in the sense that the number of computational operations of the approximation method grows at least exponentially in the problem dimension. It is the key objective of the ERC-funded MONTECARLO project to employ multilevel Monte Carlo and stochastic gradient descent type methods to design and analyse algorithms which provably overcome the curse of dimensionality in the numerical approximation of several high-dimensional functions; these include solutions of certain stochastic optimal control problems of some nonlinear partial differential equations and of certain supervised learning problems.
Objective
In a series of relevant real world problems it is of fundamental importance to approximatively compute evaluations of high-dimensional functions. Such high-dimensional approximation problems appear, e.g. in stochastic optimal control problems in operations research, e.g. in supervised learning problems, e.g. in financial engineering where partial differential equations (PDEs) and forward backward stochastic differential equations (FBSDEs) are used to approximatively price financial products, and, e.g. in nonlinear filtering problems where stochastic PDEs are used to approximatively describe the state of a given physical system with only partial information available. Standard approximation methods for such approximation problems suffer from the socalled curse of dimensionality in the sense that the number of computational operations of the approximation method grows at least exponentially in the problem dimension. It is the key objective of this project to design and analyze approximation algorithms which provably overcome the curse of dimensionality in the case of stochastic optimal control problem, nonlinear PDEs, nonlinear FBSDEs, certain SPDEs, and certain supervised learning problems. We intend to solve many of the above named approximation problems by combining different types of multilevel Monte Carlo approximation methods, in particular, multilevel Picard approximation methods, with stochastic gradient descent (SGD) optimization methods. Another chief objective of this project is to prove the conjecture that the SGD optimization method converges in the training of ANNs with ReLU activation. We expect that the outcome of this project will have a significant impact on the way how highdimensional PDEs, FBSDEs, and stochastic optimal control problems are solved in engineering and operations research and on the mathematical understanding of the training of ANNs by means of the SGD optimization methods.
Fields of science
- natural sciencescomputer and information sciencesartificial intelligencemachine learningsupervised learning
- natural sciencesmathematicsapplied mathematicsstatistics and probability
- natural sciencesmathematicspure mathematicsmathematical analysisdifferential equationspartial differential equations
- natural sciencesmathematicsapplied mathematicsnumerical analysis
Keywords
- information-based complexity
- IBC
- computational stochastics
- Monte Carlo method
- multilevel Monte Carlo method
- numerical analysis
- partial differential equation
- PDE
- backward stochastic differential equation
- BSDE
- stochastic optimal control
- stochastic partial differential equation
- SPDE
- stochastic gradient descent
- SGD
- machine learning
- artificial neural network
- ANN
Programme(s)
- HORIZON.1.1 - European Research Council (ERC) Main Programme
Funding Scheme
HORIZON-ERC - HORIZON ERC GrantsHost institution
48149 MUENSTER
Germany