Снижение точности классификатора обучения scikit после инициализации веса ⇐ Python

[Anonymous]

Я хочу реализовать базу классификатора adaboost на классификаторе sklearn, на первом этапе классификатора alg я должен инициализировать вес как «1 / # данные поезда»
но это снизит точность классификатора, я не могу понять это почему? (Я установил одинаковые веса для всех точек данных)
МОЙ КОД:
svm_weight = SVC()
svm_non_weight = SVC()

w = np.ones(len(target_train))
w.fill(float(1)/float(len(target_train)))
svm_weight.fit(data_train_feature_scaled_pca,
target_train,
sample_weight= w)

svm_non_weight.fit(data_train_feature_scaled_pca,
target_train)

print "score weight : ",svm_weight.score(data_test_feature_scaled_pca,target_train)
print "score non weight : ",svm_non_weight.score(data_test_feature_scaled_pca,target_train)

ВЫВОД:
score weight : 0.503592561285
score non weight : 0.729289940828

реализован adaboost:
class adaboost_classifier:
def __init__(self,train,target,classifier,n_estimator):
#prepairing dataset
self.N_classes = np.unique(target)
self.n_estimator = n_estimator
self.N_data = len(train)
self.trained_classifier = [[classifier,float(0),float(0), True ] for i in range(n_estimator)]
indice = []
train = np.array(train)
target = np.array(target)
dataset = np.concatenate((train,target),axis=1)
#join train and target for boosting

for i in range(len(dataset[0])-1):
indice.append(i)

self.weights = np.zeros([n_estimator,self.N_data])

# init 1/n value for weights
self.weights.fill(1/float(self.N_data))
#take sampling
new_dataset = dataset
self.N_data = len(new_dataset)
#start training sub classifier
for i in range(n_estimator):
self.loss = np.zeros(self.N_data)
#seprating training and target data
new_train = new_dataset[:,indice]
new_target = new_dataset[:,(len(dataset[0])-1)]
#train the classifier : learn f(X) with data weight
self.trained_classifier[0].fit(new_train,new_target,sample_weight=self.weights)
#computed the weighted error which is store in trained_classifier[1]
for point in range(self.N_data) :
if(self.trained_classifier[0].predict([new_train[point]]) != new_target[point]):
self.loss[point] = 1
self.trained_classifier[1] += self.weights[point]

#compute coefficient of classifier i which is store in trained_classifier[2]
self.trained_classifier[2] = 0.5 * np.log((1-self.trained_classifier[1])/self.trained_classifier[1])
#recompute weights
for j in range(self.N_data):
if(self.loss[j] == 1):
self.weights[i][j] *= np.exp(self.trained_classifier[i][2])
else:
self.weights[i][j] *= np.exp(-self.trained_classifier[i][2])

#normalizing the weights
self.trained_classifier[i][1] = self.trained_classifier[i][1] / self.weights[i].sum()


Подробнее здесь: https://stackoverflow.com/questions/418 ... ing-weight