Extensions to the OSiL schema: Matrix and cone programming Horand I. Gassmann, Dalhousie University Jun Ma, Kipp Martin, Imre Polik.

Slides:



Advertisements
Similar presentations
Optimization Services Robert Fourer, Jun Ma, Kipp Martin Optimization Services OS Server and OS Libraries Jun Ma Industrial Engineering.
Advertisements

Nonlinear Programming McCarl and Spreen Chapter 12.
Solver and modelling support for stochastic programming H.I. Gassmann, Dalhousie University Happy Birthday András, November 2009.
1 OR II GSLM Outline  some terminology  differences between LP and NLP  basic questions in NLP  gradient and Hessian  quadratic form  contour,
Automatic Control Laboratory, ETH Zürich Automatic dualization Johan Löfberg.
Convex Position Estimation in Wireless Sensor Networks
Lecture 8 – Nonlinear Programming Models Topics General formulations Local vs. global solutions Solution characteristics Convexity and convex programming.
by Rianto Adhy Sasongko Supervisor: Dr.J.C.Allwright
Graph Laplacian Regularization for Large-Scale Semidefinite Programming Kilian Weinberger et al. NIPS 2006 presented by Aggeliki Tsoli.
Continuous optimization Problems and successes
Dual Problem of Linear Program subject to Primal LP Dual LP subject to ※ All duality theorems hold and work perfectly!
1 On the Optimization of the Tay-Kingsbury 2-D Filterbank Bogdan Dumitrescu Tampere Int. Center for Signal Processing Tampere University of Technology,
Lecture outline Support vector machines. Support Vector Machines Find a linear hyperplane (decision boundary) that will separate the data.
Seminar in Advanced Machine Learning Rong Jin. Course Description  Introduction to the state-of-the-art techniques in machine learning  Focus of this.
1 Linear Programming Supplements (Optional). 2 Standard Form LP (a.k.a. First Primal Form) Strictly ≤ All x j 's are non-negative.
1 1 Slides by John Loucks St. Edward’s University Modifications by A. Asef-Vaziri.
Robert Fourer, Jun Ma, Kipp Martin, Optimization Services, January 18, 2005 Optimization Services (OS) Jun Ma Motorola, Schaumburg 01/18/2005 Robert Fourer.
How are we computing? Invalidation Certificates Consider invalidating the constraints (prior info, and N dataset units) The invalidation certificate is.
Robert Fourer, Jun Ma, Kipp Martin Optimization Services Instance Language (OSiL), Solvers, and Modeling Languages Kipp Martin University of Chicago
Optimization for Operation of Power Systems with Performance Guarantee
Optimization Services Robert Fourer, Jun Ma, Kipp Martin Setting Up and Hosting Your Solver as Web Services via Optimization Services (OS) Jun Ma
Recent Changes to the Optimization Services Project H.I. Gassmann, Faculty of Management J. Ma, Northwestern University R.K. Martin, The University of.
4.5 Solving Systems using Matrix Equations and Inverses.
4.5 Solving Systems using Matrix Equations and Inverses OBJ: To solve systems of linear equations using inverse matrices & use systems of linear equations.
General Nonlinear Programming (NLP) Software
New developments in OSiL, OSoL and OS H.I. Gassmann R.K. Martin,J. Ma INFORMS Annual meeting, Washington, DC, October 2008.
Hossein Sahebdel Iran University Of Science And Technology May 12
1 Unconstrained Optimization Objective: Find minimum of F(X) where X is a vector of design variables We may know lower and upper bounds for optimum No.
Jun Ma, Optimization Services, July 19, 2006 Optimization Services (OS) Today: open Interface for Hooking Solvers to Modeling Systems Jun Ma Northwestern.
Robert Fourer, Jun Ma, Kipp Martin Optimization Services and the Stylized “OS” Logo are registered in the US Patent & Trademark Office. All other product.
1 Kernel based data fusion Discussion of a Paper by G. Lanckriet.
Nonlinear Programming Models
An introduction to stochastic programming H.I. Gassmann.
Fourer, Ma, Martin, An Open Interface for Hooking Solvers to Modeling Systems INFORMS International, Puerto Rico, July 8-11, INFORMS International.
Robert Fourer, Jun Ma, Kipp Martin, Optimization Services, May 06, 2005 Optimization Services (OS) Jun Ma Industrial Engineering and Management Sciences.
1 Efficient Parallel Software for Large-Scale Semidefinite Programs Makoto Tokyo-Tech Katsuki Chuo University MSC Yokohama.
An XML-based schema for stochastic programs H.I. Gassmann, R. Fourer, J. Ma, R.K. Martin EURO XXI, June 2006, Reykjavik.
Introduction to Semidefinite Programs Masakazu Kojima Semidefinite Programming and Its Applications Institute for Mathematical Sciences National University.
A comparison between PROC NLP and PROC OPTMODEL Optimization Algorithm Chin Hwa Tan December 3, 2008.
Multi-area Nonlinear State Estimation using Distributed Semidefinite Programming Hao Zhu October 15, 2012 Acknowledgements: Prof. G.
OSiL: An XML-based schema for stochastic programs H.I. Gassmann, R. Fourer, J. Ma, R.K. Martin SP XI Vienna, August 2007.
A Kernel Approach for Learning From Almost Orthogonal Pattern * CIS 525 Class Presentation Professor: Slobodan Vucetic Presenter: Yilian Qin * B. Scholkopf.
Jun Ma, Sanjay Mehrotra and Huanyuan Sheng Impact Solver for Optimization Services, November 8, 2006 On Implementing a Parallel Integer Solver Using Optimization.
Support Vector Machines Reading: Ben-Hur and Weston, “A User’s Guide to Support Vector Machines” (linked from class web page)
Robert Fourer, Jun Ma, Kipp Martin Copyright 2005 Optimization Services (OS) Jun Ma Annapolis, 01/07/2005 Robert Fourer Jun Ma Northwestern University.
3.8B Solving Systems using Matrix Equations and Inverses.
Using Neumann Series to Solve Inverse Problems in Imaging Christopher Kumar Anand.
Nonlinear Dimension Reduction: Semi-Definite Embedding vs. Local Linear Embedding Li Zhang and Lin Liao.
Optimizing Your World neos &. neos & HTCondor neos : Network-Enabled Optimization System Mathematical.
Regularized Least-Squares and Convex Optimization.
Huanyuan Sheng, Sanjay Mehrotra and Jun Ma Impact Solver for Optimization Services, November 15, 2005 IMPACT Solver for Optimization Services Huanyuan(Wayne)
Parallel Direct Methods for Sparse Linear Systems
BLOM: Berkeley Library for Optimization Modeling
Solver & Optimization Problems
Boundary Element Analysis of Systems Using Interval Methods
Amir Ali Ahmadi (Princeton University)
Georgina Hall Princeton, ORFE Joint work with Amir Ali Ahmadi
Nonnegative polynomials and applications to learning
Lecture 8 – Nonlinear Programming Models
Optimization Services Instance Language (OSiL)
Local Gain Analysis of Nonlinear Systems
Quadratic Forms and Objective functions with two or more variables
Huanyuan(Wayne) Sheng
Instructors: Fei Fang (This Lecture) and Dave Touretzky
Discrepancy and Optimization
ECEN 615 Methods of Electric Power Systems Analysis
Nonlinear Dimension Reduction:
An XML-based schema for stochastic programs
Zeros of polynomial functions
Computer Animation Algorithms and Techniques
Presentation transcript:

Extensions to the OSiL schema: Matrix and cone programming Horand I. Gassmann, Dalhousie University Jun Ma, Kipp Martin, Imre Polik

© 2014 H.I. Gassmann Outline Matrices and optimization Optimization Services OSiL (Optimization Services instance Language) Cone and matrix programming Extensions to OSiL –Matrices –Cones –Matrix programming OS and CSDP

© 2014 H.I. Gassmann Matrices in optimization Coefficient matrices in linear programming Jacobian and Hessian matrices, gradient vectors Matrix variables in positive semidefinite programming Formats for sparse matrix representations –Harvard format (starts, indices and values) –MPS, CPLEX LP format –SDP format Static (real) values only, even SDP result format

© 2014 H.I. Gassmann Cone and matrix programming Constraints (and objectives) expressed in terms of cones –Second order cones –Cones of positive semidefinite matrices –Orthant cones (linear cones) –Cones of nonnegative polynomials (over some interval) Solvers: CSDP, SeDuMi, Mosek, Cplex, Gurobi, FICO Unified treatment within the Optimization Services framework

© 2014 H.I. Gassmann Sample problems Second order cone program Semidefinite program

© 2014 H.I. Gassmann CSDP Open-source project (COIN-OR) Solves Assumes A i, C, X are real and symmetric : positive semidefinite

© 2014 H.I. Gassmann Optimization Services Framework for optimization in a distributed computing environment XML schemas for communicating instances, options, results, … Implementation (COIN-OR project OS) –OSSolverService –OSAmplClient –OSServer –callable library

© 2014 H.I. Gassmann Solvers AML Corporate databases User interface Data inter- change

© 2014 H.I. Gassmann “Core” OS and OSiL Problems handled –Linear programs –Integer programs –MILP –Convex NLP –Discrete NLP –Nonconvex NLP Available solvers –Clp –Cbc –SYMPHONY –Ipopt –Bonmin –Couenne –Glpk –Cplex –Gurobi –Mosek

© 2014 H.I. Gassmann “Core” OSiL elements Variable types (C,I,B) Upper and lower bounds Possibly multi-objective Maximize or minimize Linear objective coefficients Upper and lower bounds

© 2014 H.I. Gassmann Requirements and challenges Cone inclusions, e.g., – X is symmetric, positive semidefinite – Ax lies in the rotated quadratic cone RC(1,2;3..n) Matrix expressions, e.g., – AXB + BXA –trace A T X What does a linear matrix expression look like?

© 2014 H.I. Gassmann Design principles Preserve core Be as general as possible Respect sparsity Avoid painting yourself into corners New constructs –

© 2014 H.I. Gassmann OSiL and matrices Constant matrices Matrix variables X (or ?) General matrices (e.g., Hessian, Jacobian, etc.) Symmetry Sparsity Block structure Matrix construction, e.g., A = a a T One matrix type or several?

© 2014 H.I. Gassmann OSiL: Matrix and cone extensions

© 2014 H.I. Gassmann The element

© 2014 H.I. Gassmann Example 1 – elements

© 2014 H.I. Gassmann Example 2 - transformation

© 2014 H.I. Gassmann Example 3 - blocks

© 2014 H.I. Gassmann Example 4 – base matrix <baseMatrix baseMatrixIdx="2" baseMatrixFirstRow="0" baseMatrixFirstCol="0" baseMatrixLastRow="2" baseMatrixLastCol="2" baseTranspose="false" scalarMultiplier="1.0" targetMatrixFirstRow="1" targetMatrixLastRow ="1" />

© 2014 H.I. Gassmann Example 5 – variable references

© 2014 H.I. Gassmann The element …

© 2014 H.I. Gassmann The element - Example <nonnegativePolynomialsCone maxDegree="4" numberOfRows= "2" numberOfColumns= "1"/> 2 0 Cone of symmetric positive semidefinite 2x2 matrices with nonnegative entries Cone of nonneg. poly. in two variables

© 2014 H.I. Gassmann

© 2014 H.I. Gassmann,,

© 2014 H.I. Gassmann example Recall: X = M 4 is positive semidefinite (i.e., in C 0 ): C 0 is the cone of 2x2 positive semidefinite matrices

© 2014 H.I. Gassmann example …

© 2014 H.I. Gassmann Future work Complete the parser Translate SDP problems (SDPlib) Write CSDP driver

© 2014 H.I. Gassmann How to get OS Download –Binaries –OS win32-msvc9.zipOS win32-msvc9.zip –OS linux-x86_64-gcc4.3.2.tgzOS linux-x86_64-gcc4.3.2.tgz –Stable source –OS tgzOS tgz –OS zipOS zip –Development version (using svn) svn co COIN-OS svn co COIN-OS

© 2014 H.I. Gassmann QUESTIONS?