Natural Language Processing · TEE · about-chapter

Optional extensions — after the TEE questions

Additional explanation; no corresponding question in the two supplied TEE papers. No exam marks are implied.

1. Add-k: preserve a normalized vocabulary row
2. Forward HMM: sum paths and normalize posteriors