n_redundant=7)
def svccv(C, gamma):
return cross_val_score(SVC(C=C, gamma=gamma, random_state=2),
data, target, 'f1', cv=5).mean()
def rfccv(n_estimators, min_samples_split, max_features):
return cross_val_score(RFC(n_estimators=int(n_estimators),
min_samples_split=int(min_samples_split),
max_features=min(max_features, 0.999),
random_state=2),
data, target, 'f1', cv=5).mean()
if __name__ == "__main__":
svcBO = BayesOptCV(svccv, param_grid={'C':{'type':'float', 'min':0.02, 'max':20}, 'gamma':{'type':'float', 'min':0.05, 'max':5}}, verbose=2)
svcBO.initialize(num_init=3, init_grid={'C': [0.001, 0.01, 0.1], 'gamma': [0.001, 0.01, 0.1]})
rfcBO = BayesOptCV(rfccv, param_grid={'n_estimators': {'type':'int', 'min':10, 'max':250},
'min_samples_split': {'type':'int', 'min':2, 'max':25},
'max_features': {'type':'int', 'min':0.1, 'max':0.999}}, verbose=2)
kernel_param = {} # something
acqui_param = {} # something
rfcBO.initialize(num_init=3)
svcBO.optimize(kernel_param=kernel_param, acqui_param=acqui_param)
rfcBO.optimize(kernel_param=kernel_param, acqui_param=acqui_param)
print('Final Results')
print('SVC: %f' % svcBO.report['best']['best_val'])
print('RFC: %f' % rfcBO.report['best']['best_val'])
scaler = MinMaxScaler(feature_range=(-1, 1))
train_data = scaler.fit_transform(train_data.toarray())
test_data = scaler.transform(test_data.toarray())
print train_data.shape
print test_data.shape
import numpy as np
f = lambda x: np.sin(x)
sigma = 0.02 # noise variance.
def sin(C):
return f(C) + sigma * np.random.randn(1)[0]
def svccv(C):
return cross_val_score(SVC(C=C, kernel='poly', degree=3, random_state=2),
train_data, train_labels, 'f1', cv=5).mean()
svcBO = BayesOptCV(svccv, param_grid={'C':{'type':'float', 'min':6.4e-05, 'max':60}},
bigger_is_better=True, verbose=2)
svcBO.initialize(num_init=10, init_grid={})
# kernel_param = {'nugget':0.0000001}
kernel_param = {'theta0':0.5}
acqui_param = {'kappa':2}
svcBO.optimize(kernel_param=kernel_param, acqui_param=acqui_param, n_iter=100, acqui_type='ucb', n_acqui_iter=200)
print('Final Results')
print('SVC: %f' % svcBO.report['best']['best_val'])
plt.show()
return f(C) + sigma * np.random.randn(1)[0]
def svccv(C):
return cross_val_score(SVC(C=C, kernel='poly', degree=3, random_state=2),
train_data,
train_labels,
'f1',
cv=5).mean()
svcBO = BayesOptCV(
svccv,
param_grid={'C': {
'type': 'float',
'min': 6.4e-05,
'max': 60
}},
bigger_is_better=True,
verbose=2)
svcBO.initialize(num_init=10, init_grid={})
# kernel_param = {'nugget':0.0000001}
kernel_param = {'theta0': 0.5}
acqui_param = {'kappa': 2}
svcBO.optimize(kernel_param=kernel_param,
acqui_param=acqui_param,
n_iter=100,
acqui_type='ucb',
n_acqui_iter=200)
print('Final Results')
'''
Created on Dec 18, 2015
@author: adarsh
'''
from hypop import BayesOptCV
from solutions_helper import *
import glob
trainfiles = glob.glob('/home/adarsh/Copy/SharedWorkspace/inf_ps_7/Data/train*.txt')
testfiles = glob.glob('/home/adarsh/Copy/SharedWorkspace/inf_ps_7/Data/test*.txt')
train_data, train_labels = transform_data(trainfiles)
train_data = encode_sentences(train_data)
train_size = 100
def struct_predict(C):
val_score = bpo_cost_fun(train_data[
:train_size], train_data[-500:], train_labels[:train_size], train_labels[-500:], C)
return val_score
svcBO = BayesOptCV(struct_predict, param_grid={'C':{'type':'float', 'min':6.4e-05, 'max':100}}, verbose=2)
svcBO.initialize(num_init=10, init_grid={'C': [0.001, 0.1, 1, 10, 50, 80, 100, ]})
# kernel_param = {'nugget':0.0000001}
kernel_param = {'theta0':1.0}
acqui_param = {}
svcBO.optimize(kernel_param=kernel_param, acqui_param=acqui_param, n_iter=10, n_acqui_iter=50)
print('Final Results')
print('StructSVM: %f' % svcBO.report['best']['best_val'])
plt.show()
