Skip to content

Latest commit

 

History

History
21 lines (16 loc) · 678 Bytes

README.md

File metadata and controls

21 lines (16 loc) · 678 Bytes

Baum Welch Algorithm

Implementation of Baum Welch-Algorithm for POS tagging using the expectation-minimization algorithm.

Has implementation of both online and batch training versions.

Usage

cd src/

python main_alter.py ../corpus/brown_nolines.txt

Other versions can be used depending on requirement. main_alter is the version used for calculating the results data.

Tasks

  • Implement Forward and Backward Algorithm.
  • Compute Gamma and Eta.
  • Implement Maximisation step
  • Implement scaling to avoid underflow.
  • Predict top 100 words(emission probabilities) for each tag.
  • Contains the report describing the workflow.