如何在机器学习中使用 cross_val_score 进行预测

Question

in my course i learned how to use cross validation to increase the accuracy of my model, everything looks beautiful in training. But when I go to practice in training I find that I can't use the models trained with cross-validation, follow my code:

X = array[:,0:8]
Y = array[:,8]


num_folds = 10
seed = 7


kfold = KFold(num_folds, True, random_state = seed)


modelo = LogisticRegression()


resultado = cross_val_score(modelo, X, Y, cv = kfold)


print("Acurácia: %.3f" % (resultado.mean() * 100))

in this cross-validation logic how can i use the model trained in my test data?

I'm trying something like modelo.predict(X_test) but not success

can anyone help me?

Answer 1

You need to fit the model to the data before you can use the .predict function. I believe you're using scikit learn, so:

clf = LogisticRegression()
clf = clf.fit(X, y)
clf.predict(X[:2, :])

From scikit documentation here.