A Method for Approximating Univariate Convex Functions using Only Function Value Evaluations

In this paper, piecewise-linear upper and lower bounds for univariate convex functions are derived that are only based on function value information. These upper and lower bounds can be used to approximate univariate convex functions. Furthermore, new sandwich algorithms are proposed that iteratively add new input data points in a systematic way until a desired accuracy of the approximation is obtained. We show that our new algorithms that use only function value evaluations converge quadratically under certain conditions on the derivatives. Under other conditions, linear convergence can be shown. Some numerical examples that illustrate the usefulness of the algorithm, including a strategic investment model, are given.

[1]  Lakhdar Chiter.  DIRECT algorithm: A new definition of potentially optimal hyperrectangles , 2006, Appl. Math. Comput..

[2]  David L Craft,et al.  Approximating convex pareto surfaces in multiobjective radiotherapy planning. , 2006, Medical physics.

[3]  Johannes Bisschop,et al.  AIMMS - Optimization Modeling , 2006 .

[4]  Ilya Baran.  Adaptive algorithms for problems involving black-box Lipschitz functions , 2004 .

[5]  Erik D. Demaine,et al.  Optimally Adaptive Integration of Univariate Lipschitz Functions , 2006, Algorithmica.

[6]  Pierre Hansen,et al.  On the Number of Iterations of Piyavskii's Global Optimization Algorithm , 1991, Math. Oper. Res..

[7]  Mihalis Yannakakis,et al.  Small Approximate Pareto Sets for Biobjective Shortest Paths and Other Problems , 2009, SIAM J. Comput..

[8]  Robert L. Smith,et al.  Optimal estimation of univariate black-box Lipschitz functions with upper and lower error bounds , 2003, Comput. Oper. Res..

[9]  Günter Rote,et al.  The convergence rate of the sandwich algorithm for approximating convex functions , 1992, Computing.

[10]  Alex Y. D. Siem,et al.  Derivative-free generation and interpolation of convex Pareto optimal IMRT plans , 2006, Physics in medicine and biology.

[11]  Henk Huizenga,et al.  Convex reformulation of biologically-based multi-criteria intensity-modulated radiation therapy optimization including fractionation effects , 2008, Physics in medicine and biology.

[12]  G. Rote,et al.  Sandwich approximation of univariate convex functions with an application to separable convex programming , 1991 .

[13]  C. Goh,et al.  A method for convex curve approximation , 1997 .

[14]  J. Szabados,et al.  Interpolation of Functions , 1990 .

[15]  Dick den Hertog,et al.  The Effect of Transformations on the Approximation of Univariate (Convex) Functions With Applications to Pareto Curves , 2006, Eur. J. Oper. Res..

[16]  Fernando Alonso,et al.  Intensity-modulated radiotherapy – a large scale multi-criteria programming problem , 2003, OR Spectr..

[17]  Patrice Marcotte,et al.  An optimal adaptive algorithm for the approximation of concave functions , 2006, Math. Program..

[18]  Dick den Hertog,et al.  Multivariate Convex Approximation and Least-Norm Convex Data-Smoothing , 2005, ICCSA.

[19]  Dick den Hertog,et al.  Efficient Line Searching for Convex Functions , 2004 .

[20]  David F. Pyke,et al.  Inventory management and production planning and scheduling , 1998 .

[21]  Dick den Hertog,et al.  Discrete least-norm approximation by nonnegative (trigonometric) polynomials and rational functions , 2008 .

[22]  Erik Demeulemeester,et al.  New computational results on the discrete time/cost trade-off problem in project networks , 1998, J. Oper. Res. Soc..

[23]  Pablo A. Parrilo,et al.  An Inequality for Circle Packings Proved by Semidefinite Programming , 2004, Discret. Comput. Geom..

[24]  G. Rote,et al.  Approximation of convex curves with application to the bicriterial minimum cost flow problem , 1989 .

[25]  H. Romeijn,et al.  A unifying framework for multi-criteria fluence map optimization models. , 2004, Physics in medicine and biology.

[26]  Reinaldo Morabito,et al.  AN OVERVIEW OF TRADEOFF CURVES IN MANUFACTURING SYSTEMS DESIGN , 2009 .

[27]  Edgar den Boef,et al.  Efficient Line Search Methods for Convex Functions , 2007, SIAM J. Optim..

[28]  Michel Mandjes,et al.  The Buffer-Bandwidth Trade-off Curve is Convex , 2001, Queueing Syst. Theory Appl..