Python CNN image classification model fails to classify individual images

Question

I am very new to machine learning. Looking at other work, I trained the following model https://www.kaggle.com/code/konstansan/cat-and-flower-image-classifier , which shows very high accuracy (98 - 99%) based on calculations for the testing set. To run the notebook, one needs a Kaggle account.

However, most often the model fails to classify accurately individual images from the testing set. I have 2 categories, 10 000 images per category in the training set, 2000 per category in the validation set, and 50 per category in the testing set. The images in the 3 sets are not too different in terms of size, contents, etc.

I then trained a similar model which had about 85% accuracy but still fails on individual images.

Could someone help? Thank you.

Answer 1

Sounds like model accuracy is good on training set and not really working on testing set right??, It might be because of any of the below reasons: bcz of algorithm bcz of hyper parameter bcz of data overfit bcz of data are might not be good