To summarize, a similarity measure quantifies the similarity between a pair of examples, relative to other pairs of examples. The table below compares the two types of similarity measures:
Type | Create By | Use When | Implication |
---|---|---|---|
Manual | Manually combining feature data. | Datasets are small and features are easily combined. | Gain insight into results of similarity calculations, but if feature data changes, then you must update the similarity measure. |
Supervised | Measuring distance between embeddings generated via a supervised DNN. | Datasets are large and features are hard to combine. | No insight into results, but DNN can automatically adapt to changing feature data. |