Let us see how well our model has learned from the train data, by testing it on the test data.
Note:
model.evaluate(X_test, y_test) : Returns a list with the loss(at index 0) and accuracy(at index 1) of the predictions of the model on X_test.
array[a:b] : Returns the elements of array from index a till b-1 inclusive.
model.predict : Generates output predictions for the input samples. It returns the probabilities of each class for a given test sample. We consider the classification to be of that class(or of that index) whose probability of the highest.
np.argmax : Returns the indices of the maximum values along an axis. More here.
For example,
>> a = array([[10, 11, 12],
[13, 14, 15]])
>> np.argmax(a)
5
>> np.argmax(a, axis=0)
array([1, 1, 1])
>> np.argmax(a, axis=1)
array([2, 2])
Use model.evaluate on X_test, y_test to see the test data prediction accuracy:
results = << your code comes here >>(X_test, y_test)
Print the accuracy on the test data.
print(results[1])
Let us see the predictions of 9 of the test samples from index 10.
Slice the X_test to get 9 of the data samples of the test data from index 10.
X_new = << your code comes here >>
Use model.predict on X_new and get the predictions in y_pred.
y_pred = << your code comes here >>(X_new)
Use np.argmax on y_pred along the axis=1 and print the classes predicted by the model on X_new.
print(<< your code comes here >>(y_pred, axis=1))
Let us also look at the original classes of X_new.
print (y_test[10:20])
Note - Having trouble with the assessment engine? Follow the steps listed here
Loading comments...