我在使用smote时遇到此错误

Question

enter image description here

# apply SMOTE to tackle class imbalance
from imblearn.over_sampling import SMOTE
sm = SMOTE(kind = "regular", k=1)
X_tr,y_tr = sm.fit_sample(X_train,y_train)
print(X_tr.shape)
print(y_tr.shape)

请在这件事上帮助我

Answer 1

SMOTE过采样算法没有像你建议的那样有一个名为kind的参数，就documentation而言。

自动类策略(*，

_imblearn.over_sampling.SMOTE=‘auto’，random_state=None，k_neighbors=5，n_jobs=None)

如果您指的是to、do、minority、not minority等类型的过采样，则该参数为sampling_strategy，默认为auto。

sm = SMOTE(sampling_strategy = "minority")