Beispiel #1
0
import sys
import json
from FeatFair import FeatFair
from train_models import train_feat_model
from feat_common_args import common_args

dataset = sys.argv[1]
dataset_name = dataset.split('/')[-1].split('.')[0]
attributes = sys.argv[2]
seed = int(sys.argv[3])
rdir = sys.argv[4]

common_args['obj'] = 'fitness,complexity'
common_args['pop_size'] = 100
common_args['gens'] = 0

est = FeatFair(
    **common_args,
    sel='random',
    surv='offspring',
    random_state=seed,
)

# set up Feat model
perf, hv = train_feat_model(est, 'feat_random_p100_g0', dataset, attributes,
                            seed, rdir)
Beispiel #2
0
import sys
import json
from FeatFair import FeatFair
from train_models import train_feat_model
from feat_common_args import common_args

dataset = sys.argv[1]
dataset_name = dataset.split('/')[-1].split('.')[0]
attributes = sys.argv[2]
seed = int(sys.argv[3])
rdir = sys.argv[4]

common_args['obj'] = 'fitness'
est = FeatFair(
    **common_args,
    sel='tournament',
    surv='offspring',
    random_state=seed,
)

# set up Feat model
perf, hv = train_feat_model(est, 'feat_tourn', dataset, attributes, seed, rdir)
Beispiel #3
0
import sys
from train_models import train_feat_model
from feat_common_args import common_args
from FeatFair import FeatFair
import json

dataset = sys.argv[1]
dataset_name = dataset.split('/')[-1].split('.')[0]
attributes = sys.argv[2]
seed = int(sys.argv[3])
rdir = sys.argv[4]

est = FeatFair(
            **common_args,
            sel = 'fair_lexicase2',
            surv = 'nsga2',
            random_state=seed,
            )

# set up Feat NSGA2 model
perf, hv = train_feat_model(est, 'feat_flex2_nsga2', dataset, attributes, seed,
        rdir)