home
Home
Modeling Languages
    ModelingToolkitCatalystNBodySimulatorParameterizedFunctionsProcessSimulatorMomentClosure
Model Libraries and Importers
    ModelingToolkitStandardLibraryModelingToolkitNeuralNetsDiffEqCallbacksFiniteStateProjectionCellMLToolkitSBMLToolkitBaseModelicaAudioPluginsReactionNetworkImportersDiffEqPhysicsDiffEqFinancialPubChemPyomoMathML
Symbolic Tools
    ModelOrderReductionSymbolicsSymbolicUtilsSymbolicIntegrationSymbolicSMTSymbolicLimitsSymbolicAnalysisFunctionProperties
Array Libraries
    RecursiveArrayToolsComponentArraysLabelledArraysMultiScaleArrays
Equation Solvers
    LinearSolveNonlinearSolveDifferentialEquationsIntegralsDifferenceEquationsOptimizationJumpProcessesLineSearchEvolutionaryNeuralLinearSolveCorleone
Inverse Problems / Estimation
    CurveFitSciMLSensitivityDiffEqParamEstimDiffEqBayes
PDE Solvers
    MethodOfLinesNeuralPDENeuralOperatorsFEniCSHighDimPDEDiffEqOperatorsFiniteVolumeMethodFiniteVolumeMethod1DPDEBase
Advanced Solver APIs
    OrdinaryDiffEqBoundaryValueDiffEqDiffEqGPUSteadyStateDiffEqOrdinaryDiffEqOperatorSplittingIRKGaussLegendreMATLABDiffEqPETScDiffEqQuantumNLDiffEq
Parameter Analysis
    EasyModelAnalysisGlobalSensitivityStructuralIdentifiabilityMinimallyDisruptiveCurvesCatalystNetworkAnalysis
Third-Party Parameter Analysis
    BifurcationKit
Uncertainty Quantification
    PolyChaosSciMLExpectationsOptimalUncertaintyQuantification
Function Approximation
    SurrogatesReservoirComputing
Implicit Layer Deep Learning
    DiffEqFluxDeepEquilibriumNetworksNeuralLyapunov
Symbolic Learning
    DataDrivenDiffEqSymbolicNumericIntegration
Third-Party Differentiation Tooling
    SparseDiffToolsFiniteDiff
Numerical Utilities
    ExponentialUtilitiesDiffEqNoiseProcessPreallocationToolsEllipsisNotationDataInterpolationsDataInterpolationsNDPoissonRandomQuasiMonteCarloRuntimeGeneratedFunctionsMuladdMacroFindFirstFunctionsSparseDiffToolsBipartiteGraphsFastAlmostBandedMatricesFastBroadcastFunctionWrappersWrappersLHLFactorizationLightweightStatsPureGebalPureKLUPureUMFPACKRootedTreesSparseBandedMatricesRespecializeParamsConcreteStructs
High-Level Interfaces
    SciMLBaseSciMLStructuresSciMLLoggingADTypesSymbolicIndexingInterfaceTermInterfaceSciMLOperatorsSurrogatesBaseCommonSolveSciMLIteratorsStatic
Third-Party Interfaces
    ArrayInterfaceStaticArrayInterface
Developer Documentation
    SciMLStyleColPracDiffEq Developer DocumentationOrgMaintenanceScripts
Extra Resources
    SciMLWorkshopExtended SciML TutorialsThe SciML BenchmarksModelingToolkitCourse
Dyad by JuliaHub
    DyadBY JULIAHUB

    Dyad by JuliaHub is the agentic modeling and simulation platform built on the SciML stack.

    • Component-based GUI for modeling
    • Modeling language and compiler stack integrated with and transparent to AI agents
    Explore Dyad ↗

    JuliaHub offers commercial support for ModelingToolkit and the SciML ecosystem. Contact Us.

Products built with SciML
    • Dyad
    • Pumas
    • Cedar EDA
    • Neuroblox
    • Planting Space
    /
    HighDimPDE.jl logo
    HighDimPDE.jl
    • Home
    • Getting started
    • Problems
    • SciMLBase Interface
    • Public API
    • Solver Algorithms
      • The MLP algorithm
      • The DeepSplitting algorithm
      • The DeepBSDE algorithm
        • The general idea 💡
        • References
      • The NNStopping algorithm
      • The NNKolmogorov algorithm
      • The NNParamKolmogorov algorithm
    • Tutorials
      • Solving the 10-dimensional Fisher-KPP equation with DeepSplitting
      • Solving a 100-dimensional Hamilton-Jacobi-Bellman Equation with DeepBSDE
      • Solving the 10-dimensional Fisher-KPP equation with MLP
      • NNStopping
      • NNKolmogorov
      • NNParamKolmogorov
    • Extended Examples
      • Solving the 100-dimensional Black-Scholes-Barenblatt Equation
    • Solver Algorithms
      • The MLP algorithm
      • The DeepSplitting algorithm
      • The DeepBSDE algorithm
    • Feynman Kac formula
    Version
    • Solver Algorithms
    • The DeepBSDE algorithm
    • The DeepBSDE algorithm
    GitHub

    The DeepBSDE algorithm

    Problems Supported:

    1. ParabolicPDEProblem

    The general idea 💡

    The DeepBSDE algorithm is similar in essence to the DeepSplitting algorithm, with the difference that it uses two neural networks to approximate both the the solution and its gradient.

    References

    • Han, J., Jentzen, A., E, W., Solving high-dimensional partial differential equations using deep learning. arXiv (2018)
    « The DeepSplitting algorithmThe NNStopping algorithm »

    Powered by Documenter.jl and the Julia Programming Language.

    Settings


    This document was generated with Documenter.jl version 1.19.0 on Saturday 17 October 2026. Using Julia version 1.10.12.