Skip to content

Source for "Multi-Objective Evolutionary Optimisation of Antibiotic Treatments" (Ochoa, Christie, Brownlee, & Hoyle, 2020).

Notifications You must be signed in to change notification settings

leechristie/antibiotic-optimisation

Repository files navigation

Antibiotic Optimisation

Model (antibiotic-model)

A stochastic model for antibiotic resistance.

Click to see documentation...

Example Usage (Model)

AntibioticModel can be used directly, giving results to compare to the MATLAB reference implementation. The model constructor accepts additional parameters to specify initial bacterial load and random number generator for the simulation.

int samples = 1000;
AntibioticModel model = new AntibioticModel(samples);

int[] solution = new int[] {10, 10, 10, 10, 10, 10, 10, 10, 10, 10};

double fitness = model.evaluate(solution);

Example Usage (Problem)

AntibioticProblem is used to create an IntegerProblem instance for jMetal algorithms. Many objectives may be specified.

int maxIndividualDosage = 60;
int maxConcentraition = 60;
AntibioticProblem problem = new AntibioticProblem(
        model,
        maxIndividualDosage,
        AntibioticObjective.totalAntibiotic(),
        AntibioticObjective.overdoseAmount(maxConcentraition),
        AntibioticObjective.uncuredProportion());

Algorithm<List<DoubleSolution>> algorithm = ... // jMetal Algorithm usage

Experiments (antibiotic-model-experiments)

Documentation to follow

Offline Re-evaluation (antibiotic-offline-reeval)

Documentation to follow

Results Post-Processing (results-post-processing)

Documentation to follow