Set-Based Differential Evolution Algorithm Based on Guided Local Exploration for Automated Process Discovery

Algorithm 1

Pseudocode of DEMiner.

(1)	Initialize population
(2)	Evaluate population
(3)	Calculate meanFitness and devFitness of the population
(4)	generation ⟵ 1, timesNotChange ⟵ 0
(5)	while generation ≤ maxGenerations && timesNotChange ≤ maxNotChange do
(6)	if meanFitness ≥ MF && devFitness ≤ DF && rand ≤ R do
(7)	Generate the trial individuals by the guided local exploration
(8)	else
(9)	Generate the trial individuals by the DE algorithm
(10)	Evaluate the trial individuals
(11)	if the fitness of the trial individuals is higher that the fitness of the target targets do
(12)	Replace population
(13)	timesNotChange ⟵ 0
(14)	else
(15)	timesNotChange++
(16)	Update meanFitness and devFitness
(17)	generation++