Deep Learning Illustrated: A Visual, Interactive Guide to Artificial Intelligence

Tables

2.1     Contrasting attributes of localist, one-hot representations of words with distributed, vector-based representations

2.2     Traditional machine learning and deep learning representations, by natural language element

8.1     Cross-entropy costs associated with selected example inputs

11.1   Comparison of word2vec architectures

11.2   The words most similar to select test words from our Project Gutenberg vocabulary

11.3   A confusion matrix

11.4   Four hot dog / not hot dog predictions

11.5   Four hot dog / not hot dog predictions, now with intermediate ROC AUC calculations

11.6   Comparison of the performance of our sentiment classifier model architectures

14.1   Fashion-MNIST categories