I am trying to develop a model with three classes. To do so, I tried to develop a model with different combinations of the data samples in each class.

For example: the $1^{st}$ model has 500 images each in each of the three classes. For the $2^{nd}$ I developed three models with 250 images in one class and 500 images in the other two classes (in this combination).

(No change for the data samples in the test folder for every model).

So I checked the accuracy and precision of each of the model. My doubt is, is this imbalance in the data samples affecting the overall precision of the model with significantly large changes or not?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.