knn = KNeighborsClassifier(n_neighbors=9)
knn.fit(X_train, y_train)
y_pred = knn.predict(X_test)
print(metrics.accuracy_score(y_test, y_pred))

# Output
0.9736842105263158