0
$\begingroup$

I am creating a sentiment analysis model using Naive Bayes. When I test the model, I get an average accuracy of 65%; however, the sensitivity of the model is much higher, 90%.

So, I am wondering if there are methods to fixing this data; or, since the sensitivity is very high, then would it be ok to move forward with the model?

Improve this question
$\endgroup$
0

1 Answer 1

Reset to default
1
$\begingroup$

I could get perfect sensitivity for positive sentiment if I always predict positive sentiment, but my accuracy could be 50%ish depending on the distribution of positive sentiment in the data. The sensitivity and accuracy scores alone are not enough to tell you if your model is any good, you will need to have some goal that you are trying to achieve e.g., get 70% accuracy.

Improve this answer
$\endgroup$
1
  • $\begingroup$ Additionally to what Cameron said, check for bias in your dataset, if your dataset is 65% positive... I think you can probably see why that would be an issue $\endgroup$
    – Recessive
    Commented Dec 8, 2021 at 4:12

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .