How to get the prediction value in LSTM model? - deep-learning

Question

This model is trying to classify the emails whether they are spam or ham.
The training is finished, but how should I input the new email text and get the value that can tell me if the email is spam or ham ?
This is the example code in the text book:
import sklearn
from sklearn.metrics import precision_recall_fscore_support as score
# prediction on test data
predicted_blstm=model.predict(test_data)
predicted_blstm
# model evaluation
from sklearn.metrics import precision_recall_fscore_support as score
precision, recall, fscore, support = score(labels_test, predicted_blstm.round())
print('precision: {}'.format(precision))
print('recall: {}'.format(recall))
print('fscore: {}'.format(fscore))
print('support: {}'.format(support))
print("############################")
print(sklearn.metrics.classification_report(labels_test, predicted_blstm.round()))
and the result:
precision: [0.98782961 0.95348837]
recall: [0.99387755 0.91111111]
fscore: [0.99084435 0.93181818]
support: [980 135]
############################
precision recall f1-score support
0 0.99 0.99 0.99 980
1 0.95 0.91 0.93 135
micro avg 0.98 0.98 0.98 1115
macro avg 0.97 0.95 0.96 1115
weighted avg 0.98 0.98 0.98 1115
samples avg 0.98 0.98 0.98 1115
I had tested the code
mode.predict()
but the result is like:
array([[9.9973804e-01, 2.6198191e-04],
[9.9988401e-01, 1.1600493e-04],
[9.9996233e-01, 3.7628190e-05],
[9.9998081e-01, 1.9162568e-05],
[9.9998498e-01, 1.5043216e-05],
[9.9907982e-01, 9.2014833e-04],
...
[9.9996233e-01, 3.7628190e-05],
[9.9996233e-01, 3.7628190e-05]], dtype=float32)
What does this number means ?
Can I get the answer from this array by showing the message "spam" or "ham" ?

Answer 1

I don't know the features in your train and test data, but if your model is trained only on the email text feature, then you could the following:
1) Convert the email text, for example "This is the email I want to test", into a vector using the vectorizer used on the training data.
2) If your vector is stored in a variable 'vec'. Then you can predict if the email is ham or spam using
prediction = model.predict(vec)
The varaible 'prediction' will hold your answer.