Distant Supervision with Transductive Learning for Adverse Drug Reaction Identification from Electronic Medical Records

Algorithm 1

Pseudocode for EM with NB-independent assumption (iEM).

 Input:

  = the number of labels

  = the maximum number of iteration

 Output: parameter

1

2 ;

3 repeat

4 for to n do

5   E–step:

     Estimate model probability:   (9)

     M–step:

     Update class-conditional probability:   (10)

     Update class probability:   (11)

6 

7 until convergence or