Regarding training and validation accuracy
1 次查看(过去 30 天)
显示 更早的评论
I have trained the CNN model for dataset of 220 images and attained the results as shown below:
![](https://www.mathworks.com/matlabcentral/answers/uploaded_files/694004/image.png)
Is this network overfit or not?? Can I consider these results genuine for further study? What is difference between mini-batch loss and validation loss? what is perfect size of mini-batch for these number of images? Can CNN be applied to small image datatsets?
1 个评论
KSSV
2021-7-24
To find out net is overfit or not, test it with the known data. Rest of the definitions are given at the doc.
回答(0 个)
另请参阅
类别
在 Help Center 和 File Exchange 中查找有关 Image Data Workflows 的更多信息
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!