December 2019
Intermediate to advanced
468 pages
14h 28m
English
CBOW predicts the most likely word given its context (surrounding words). For example, given the sequence The quick _____ fox jumps, the model will predict brown. The context is the n preceding and the n following words of the word in focus (unlike NPLM, where only the preceding words participate). The following screenshot shows the context window as it slides across the text:

CBOW takes all words within the context with equal weights and doesn't consider their order (hence the bag in the name). It is somewhat similar to NPLM, ...
Read now
Unlock full access