Mastering Python for Finance - Second Edition

Running risk metrics

In Chapter 10, Machine Learning for Finance, we discussed the confusion matrix, accuracy score, precision score, recall score, and F1 score in measuring classification-based predictions. We can reuse those metrics on our model as well.

Since the model output is in the normalized decimal format between 0 and 1, we round it up to the nearest 0 or 1 integer to obtain the predicted binary classification labels:

In [ ]:    predictions = model.predict(test_scaled_x)    pred_values = predictions.round().ravel()

The ravel() command presents the result as a single list stored in the pred_values variable.

Compute and display the confusion matrix:

In [ ]:    from sklearn.metrics import confusion_matrix matrix = confusion_matrix(test_y, ...