42-AI/bootcamp_machine-learning

ML03 ex03 epsilon should be added to y_hat, not y

stelon77 opened this issue · 0 comments

  • Day: 03
  • Exercise: 03

'The purpose of epsilon (eps) is to avoid log(0) errors, it is a very
small residual value we add to y. '
should be :
The purpose of epsilon (eps) is to avoid log(0) errors, it is a very
small residual value we add to y_hat

as only h_hat is submitted to log errors

Fixed on:

  • Github
  • Gitlab