Accuracy Vs F1-score - Githubissues

chakki-works / seqeval

A Python framework for sequence labeling evaluation(named-entity recognition, pos tagging, etc...)

MIT License

1.09k stars 129 forks source link

Accuracy Vs F1-score #83

Open yassmine-lam opened 3 years ago

yassmine-lam commented 3 years ago

Hi,

I am working on a sequence labeling problem. I am using seqeval to obtain accuracy and F1 scores. However, is it normal to get an accuracy significantly higher than F1-score (Acc 95.2% and average F-1 score : 52.4%)

I am very confused

Thank u for ur help

wzkariampuzha commented 3 years ago

Yes, see here: https://towardsdatascience.com/accuracy-precision-recall-or-f1-331fb37c5cb9