A Diverse Stochastic Search Algorithm for Combination Therapeutics
Figure 1
Minimum distance to optimal points for function f4a. We simulate the f4a function (see Supplementary Table 6) 100 times with two different LHC numbers. First figure represents the simulation with LHC equal to 40 and second figure represents the simulation LHC equal to 10. The analyzed function has 1 optimal point. The blue line represents the minimum norm 1 distance between the optimal points and DSS selected points. The red vertical line represents the end of Step 1, that is, Latin Hypercube Numbers, which is equal to 40th iteration in first simulation set and 10th iteration in the second simulation set; and the black vertical dotted line represents the average value of iterations (cost of proposed algorithm) required to find one of the points with (equal to 79.25). The cyan vertical dotted line represents the worst situation out of 100 different runs.
