def nearmiss1(X, y):
undersample = NearMiss(version=1, n_neighbors=3)
X, y = undersample.fit_resample(X, y)
counter = Counter(y)
print("NearMiss-1", counter)
plot_dataset(X, y, counter)
def weighted_neural_network(n_input):
model = define_model(n_input)
weights = {0: 1, 1: 100}
history = model.fit(trainX,
trainy,
class_weight=weights,
epochs=100,
verbose=0)
yhat = model.predict(testX)
score = roc_auc_score(testy, yhat)
print('{0:<30} ROC AUC: {1:.3f}'.format('Weighted Neural Network', score))
plot_dataset(testX, array([round(i[0]) for i in yhat]),
Counter(array([round(i[0]) for i in yhat])))
# borderline-SMOTE for imbalanced dataset
from collections import Counter
from sklearn.datasets import make_classification
from imblearn.over_sampling import BorderlineSMOTE
from plotDataset import plot_dataset
X, y = make_classification(n_samples=10000,
n_features=2,
n_redundant=0,
n_clusters_per_class=1,
weights=[0.99],
flip_y=0,
random_state=1)
counter = Counter(y)
print(counter)
plot_dataset(X, y, counter)
oversample = BorderlineSMOTE()
X, y = oversample.fit_resample(X, y)
counter = Counter(y)
print(counter)
plot_dataset(X, y, counter)
score = roc_auc_score(testy, yhat)
print('{0:<30} ROC AUC: {1:.3f}'.format('Standard Neural Network', score))
def weighted_neural_network(n_input):
model = define_model(n_input)
weights = {0: 1, 1: 100}
history = model.fit(trainX,
trainy,
class_weight=weights,
epochs=100,
verbose=0)
yhat = model.predict(testX)
score = roc_auc_score(testy, yhat)
print('{0:<30} ROC AUC: {1:.3f}'.format('Weighted Neural Network', score))
plot_dataset(testX, array([round(i[0]) for i in yhat]),
Counter(array([round(i[0]) for i in yhat])))
if __name__ == '__main__':
trainX, trainy, testX, testy = prepare_data()
n_input = trainX.shape[1]
plot_dataset(testX, testy, Counter(testy))
standard_neural_network(n_input)
weighted_neural_network(n_input)