AD2016 - The 7th International Conference on
Algorithmic Differentiation
Programme and Presentations

Conference Chairs:
Patrick Farrell (University of Oxford, UK)
Shaun Forth (Cranfield University, UK)
Andreas Griewank (Yachay Tech, Ecuador)
Paul Hovland (Argonne National Laboratory, USA)
Jens-Dominik Müller (Queen Mary University of London, UK)

Monday 12th - Thursday 15th September 2016
Christ Church Oxford, UK

Monday 12th September 2016

08:30Registration and Opening
 Piecewise linear and non-smooth I - Chair Andrea Walther
08:50Kamil KhanBranch-locking AD techniques for composite nonsmooth functions and nonsmooth implicit functions
09:10Richard HasenfelderGeneralized Mid-Point and Trapezoidal Rules for Lipschitzian RHS
09:30Tom StreubelNewton-like Methods for Solving Piecewise Smooth Systems of Equations Based on Successive Piecewise Tangent or Secant Linearization
09:50Break
 Invited Presentation - Chair Andreas Griewank
10:20Siegfried M. RumpReliable Computing and AD in Matlab/Octave
 Piecewise linear and non-smooth II - Chair Andreas Griewank
11:20Sabrina FiegeA Method for the Minimization of Piecewise Smooth Objective Functions
11:40Koichi KubotaSimple implementation and examples for piecewise linearization with abs-normal form
12:00Poster Presentations I - Chair Shaun Forth
 Laurent HascoëtPiecewise Linear AD via Source Transformation
 Manuel RadonsPerturbation-Stable Surjectivity of Piecewise Affine Functions
 Jean UtkeA Nonsmoooth Optimization Case Study
 Benjamin JurgelucksParallel Algorithmic Differentiation in OpenModelica
12:20Lunch in the Great Hall
 Invited Presentation - Chair Paul Hovland
13:50Paul I. BartonComputationally Relevant Generalized Derivatives: Theory, Evaluation and Applications
 Parallelisation and Benchmarking I - Chair Paul Hovland
14:50Markus TowaraSIMPLE Adjoint Message Passing
15:10Daniel WaltherTransforming OpenMP Programs by Automatic Differentiation: Beyond Preserving the Parallelization Specified in the Original Program
15:30Break
 Parallelisation and Benchmarking II - Chair Patrick Farrell
16:00Jan HückelheimExact replication of OpenMP-parallelism in reverse-mode AD for loops with symmetric memory access
16:20Mahesh NarayanamurthiAD-Suite - A Test Suite for Algorithmic Differentiation
16:40Filip SrajerA Benchmark of Selected Algorithmic Differentiation Tools on Some Problems in Machine Learning and Computer Vision
17:00Shaun ForthA Matlab implementation of the Minpack-2 Test Problem Collection
17:20Close

Tuesday 13th September 2016

 And now for something completely ... - Chair Chris Bischof
08:30Bradley BellThe Newton Step Method for Algorithmic Differentiation with Implicit Functions
08:50Toshio FukushimaAlgorithmic Computation of Finite Difference
09:10Barak A PearlmutterTricks from Deep Learning
09:30Paul HovlandA Forward Mode Version of the CENA Method for Roundoff Error Estimation
09:50Break
 Invited Presentation - Chair Laurent Hascoët
10:20Daniel GoldbergAutomatic differentiation of current ice sheet models
 Iterative Systems - Chair Laurent Hascoët
11:20Bruce ChristiansonDifferentiating through Conjugate Gradient
11:40Siamak AkbarzadehConsistent treatment of incompletely converged iterative linear solvers in reverse-mode AD
12:00Poster Presentations II - Chair Jens-Dominik Müller
 Jonathan HüserIntegration of Differential-Algebraic Equations with Optimality Criteria
 Johannes LotzMixed Integer Programming for Call Tree Reversal
 Barak A. PearlmutterEvolving the Incremental Lambda Calculus Into a Model of Forward AD
12:20Lunch in the Great Hall
 AD for Other Languages and Interoperability I - Chair Jean Utke
13:50Kshitij KulshreshthaInterfacing source transformation AD with operator overloading libraries
14:10Dominic JonesBlock Scope Differentiation
14:30Valérie PascualMixed-language Automatic Differentiation
14:50Laurent HascoëtInterfacing OpenAD and Tapenade
15:10Jarrett RevelsJulia for Automatic Differentiation
15:30Break
 AD for Other Languages and Interoperability II - Chair Valérie Pascual
16:00Jeffrey Mark SiskindEfficient Implementation of a Higher-Order Language with Built-In AD
16:20Feng QiangOn efficient Hessian computation using the edge pushing algorithm in Julia
16:40Atilim Güneş BaydinDiffSharp: An AD Library for .NET Languages
17:00Kshitij KulshreshthaWrappers for ADOL-C in scripting languages using SWIG
17:20Close

Wednesday 14th September 2016

 Exploiting Structure I - Chair Bruce Christianson
08:30Alex PothenEnhancing sparsity of Hermite polynomial expansions with Automatic Differentiation
08:50Uwe NaumannAdjoint Code Design Patterns
09:10Jeffrey Mark SiskindBinomial Checkpointing for Arbitrary Programs with No User Annotation
09:30Viktor MosenkisOn Lower Bounds for the Optimal Jacobian Accumulation Problem on linearized DAGs
09:50Break
 Invited Presentation - Chair Andreas Griewank
10:20Andrea WaltherCovering resilience: A recent development for binomial checkpointing
 Exploiting Structure II - Chair Andreas Griewank
11:20Martin BückerFirst-order Derivatives of Associated Legendre Functions
11:40Mohammad Ali RostamiBidirectional Partitioning is no Better than Unidirectional Partitioning when Computing the Diagonal Entries of a Sparse Jacobian Matrix
12:00Poster Presentations III- Chair Jens-Dominik Müller
 Alexander HückA Case Study of ADOL-C and CoDiPack applied to the Ice Sheet System Model
 Guillaume SalVibrato and Automatic Differentiation for High Order Derivatives and Sensitivities of Financial Options
 Gilberto PèrezAn Approach to Computing Discrete Adjoints for MPI-Parallelized Models Applied to the Ice Sheet System Model
12:20Lunch in the Great Hall
 Invited Presentation - Chair Uwe Naumann
13:50Jacques du ToitApplying AD in the Finance Industry, and an Operator Overloading Tool for “Handwritten” Adjoints
 Exploiting Structure III- Chair Uwe Naumann
14:50John PryceDifferential-Algebraic Equations, the Dummy Derivatives Method, and AD
15:10Alexander HückOperator Overloading Compatibility for AD - A Case Study of Scientific C++ Codes
15:30Break + Wednesday Posters
 C++ - Chair Martin Bücker
16:20Eric PhippsOperator Overloading-based Automatic Differentiation of C++ Codes on Emerging Manycore Architectures
16:40Max SagebaumExpression templates for primal value taping in the reverse mode of AD
17:00Laurent HascöetSource-transformation Differentiation of a C++-like Ice Sheet model
17:20Close
18:15Drinks in the Masters Garden
19:00Conference Dinner in the Great Hall

Thursday 15th September 2016

 Applications - Chair Ralf Giering
08:30Mladen BanovicAutomatic Differentiation of a CAD System applied in an Adjoint CFD Method
08:50Lisa KuschSemi-automatic transition from simulation to one-shot optimization with equality constraints
09:10Marc SchwalbachDevelopments of a Discrete Adjoint Structural Solver for Shape and Composite Material Optimization
09:30Benjamin JurgelucksOptimisation of triple-ring-electrodes on piezoceramic transducers using algorithmic differentiation
09:50Break
 Invited Presentation - Chair Paul Hovland
10:20Patrick FarrellAutomatic parallel adjoints of finite element simulations with dolfin-adjoint
 Higher Order Derivatives I- Chair Paul Hovland
11:20Jens GustedtArbogast - Higher order AD for special functions with Modular C
11:40Mu WangSecond Order Reverse Mode of AD - A Live Variable Approach for Evaluating Hessians
12:00Poster Presentations IV- - Chair Jens-Dominik Müller
 Viktor MosenkisAdjoint NAG Library
 Arindam SenDiscrete adjoint OpenFoam for surface and volume sensitivities
 Olivier MullierValidated Computation of the Local Truncation Error of Runge-Kutta Methods with Automatic Differentiation
12:20Lunch in the Great Hall
 Higher Order Derivatives II - Chair Andrea Walther
13:50Richard NeidingerComparing High-Order Multivariate AD Methods
14:10Shahadat HossainPattern Graph for Sparse Hessian Matrix Determination
14:30Mu WangHigh Order Reverse Mode of AD - Theory and Implementation
14:50Isabelle CharpentierA full higher-order AD continuation and bifurcation framework
15:10Closing Remarks
15:30End of Conference

This document was translated from LATEX by HEVEA.