Documentation of 'ec.app.bbob.BBOBenchmarks' Java class
BBOBenchmarks
ec.app.bbob

Class BBOBenchmarks

  • All Implemented Interfaces:
    Prototype, Setup, SimpleProblemForm, java.io.Serializable, java.lang.Cloneable


    public class BBOBenchmarks
    extends Problem
    implements SimpleProblemForm
    The Black Box Optimization workshop (BBOB) has an annual competition for doing real-valued parameter optimization. The examples shown here are more or less faithful reproductions of the BBOB 2010 C code, only using Mersenne Twister instead of BBOB's random number generator. Unfortunately, the original BBOB code has various magic numbers, unexplained variables, and unfortunate algorithmic decisions. We've reproduced them exactly rather than attempt to convert to a standard ECJ template, and simply apologize beforehand.

    Parameters

    base.type
    String = none (default) , sphere, ellipsoidal, rastrigin, buch-rastrigin, linear-slope, attractive-sector, step-elipsoidal, rosenbrock, rosenbrock-rotated, ellipsoidal-2, discus, bent-cigar, sharp-ridge, different-powers, rastrigin-2, weierstrass, schaffers-f7, schaffers-f7-2, griewak-rosenbrock, schwefel, gallagher-gaussian-101me, gallagher-gaussian-21hi, katsuura, lunacek
    (The particular function)
    base.noise
    String = none (default) , gauss, uniform, cauchy, gauss-moderate, uniform-moderate, cauchy-moderate
    (what type of noise (if any) to add to the function value)
    base.reevaluate-noisy-problems
    boolean = true(default)
    (whether to reevaluate noisy problems)
    See Also:
    Serialized Form

DMelt 3.0 © DataMelt by jWork.ORG

You see the box below because you did not login.