pso {automl}R Documentation

PSO parameters and hyperparameters

Description

List of parameters and hyperparameters for Particle Swarm Optimization

Arguments

All PSO parameters and hyperparameters are preset with default value

psopartpopsize

number of particles in swarm (discrete value)
(‘autopar’ context: default value 8, which means that 8 different neural net hyperparameters sets will be tested
(‘hpar’ context: default value 50, which means that 50 neural net weights sets will be tested
#tuning priority 1

(impact on memory consumption)

CAUTION: you should only change the values below if you know what you are doing

psovarvalmin

Minimum value for particles positions (default value -10)

psovarvalmax

maximum value for particles positions (default value 10)

psovelocitymaxratio

ratio applied to limit velocities (continuous value between 0 and 1, default value 0.2)

psoinertiadampratio

inertia damp ratio (continuous value between 0 and 1, default value 1 equivalent to OFF)

psokappa

kappa (default value 1)

psophi1

Phi 1 (default value 2.05)

psophi2

Phi 2 (default value 2.05)


back to autopar, hpar

See Also

PSO video tutorial from Yarpiz


[Package automl version 1.3.2 Index]