HDOSE description 🔗

The High Dimension Origin Space Exploration (HDOSE) method is used to explore the multiples antecedents of a pattern when the input space is of high dimension (i.e. > 6). It produces input parameter values that produce a given pattern described by a set of constraints on the objectives (i.e. objectives should be under given thresholds). HDOSE optimize the fitness and when it finds solutions that are good enough it keep them and archive. New solutions are added to the archive when there input values are distant of at least a distance d = 1.0 of all existing solution in the archive (the distance L1 is used). When the archive size reaches 500 a new value for d is computed to shrink the archive size under 500 given that all the solutions present in the archive are distant of at least d. The hook arguments for the OSEEvolution are:

• output: the file in which to store the result,
• keepHistory: optional, Boolean, keep the history of the results for future analysis,
• frequency: optional, Long, the frequency in generations where the result should be saved, it is generally set to avoid using too much disk space,
• keepAll: optional, Boolean, save all the individuals of the population not only the optimal ones.

Example 🔗

Here is a use example of the OSE method in an OpenMOLE script:

val myseed = Val[Long]

val param1 = Val[Array[Double]]
val param2 = Val[Double]
val param3 = Val[Array[Boolean]]
val param4 = Val[Array[Int]]

val output1 = Val[Double]
val output2 = Val[Double]

HDOSEEvolution(
  evaluation = modelTask,
  parallelism = 10,
  termination = 100,
  origin = Seq(
    param1 in Seq.fill(10)(0.0 to 1.0),
    param2 in (-10.0 to 10.0 weight 10.0),
    param3 in Seq.fill(5)(TrueFalse weight 2.0),
    param4 in Seq.fill(10)(0 to 100)
  ),
  objective = Seq(
    output1 under 5.0,
    output2 under 50.0),
  stochastic = Stochastic(seed = myseed)
) hook (workDirectory / "results.omr", frequency = 100)

origin describes the genome of the optimisation. To compute the distance, each part of the genome are normalised. If you want to give more weight to a given part of the genome in the distance computation you can use the weight keyword.
objective describe the pattern to reach with inequalities. The patten is considered as reached when all the objective are under their threshold value. In this example HDOSE computes a maximal diversity of inputs for which all the outputs are under their respective threshold values.