Kazjon/GloVex

Issues

• Check that combine_surprise_normalised isn't cheatig by using the observed rather than calculated surprises.

  #13 opened 7 years ago
  0

• no_below parameter in evaluate_personalised does not appear to be working?

  #11 opened 7 years ago
  2

• evaluate_personalised crashes with keyerror: "italian"

  #10 opened 7 years ago
  4

• "combined" surprise values appear to be too high in some cases

  #9 opened 7 years ago
  1

• Wrong indexing of the dictionary surps in the document_surprise function of the <evaluate_personalized.py>

  #7 opened 7 years ago
  4

• The <evaluate.py> script is printing the least surprising (high scores) documents/recipes instead of the most surprising

  #6 opened 7 years ago
  3

• High surprising scores per cuisine when using fam_cat <preprocessor.py>

  #5 opened 7 years ago
  2

• Title

  #4 opened 7 years ago
  0