Accuracy doesnot change by changing number of layers and filters.

2 次查看（过去 30 天）
Saira charan
Saira charan 2018-7-31
回答： BERGHOUT Tarek 2019-2-4
I am using cnn for my dataset which includes 322 grayscale images of two different sizes, 1024*1024 and 224*224. No matter how many layers I increase or decrease in my code it doesnot change it's accuracy, as it remains the same 64%. I think there is some issue in my code but it runs perfectly whenever i make changes to it. Can anybody suggest what might be wrong?
  1 个评论
shefali saxena
shefali saxena 2019-1-19
hello
I am facing same problem while training a ECG signals.
my mini batch loss is fixed at 0.69xx and accuracy at 50%.
how to increase accuracy ??
plregardsase help!!!
Thanks and regards

请先登录，再进行评论。

请先登录，再回答此问题。

回答（1 个）

BERGHOUT Tarek
BERGHOUT Tarek 2019-2-4
try toscale your data between (0 and 1) even thje targets and try a gain , because i faced those problems too.

类别

在 Help Center 和 File Exchange 中查找有关 Recognition, Object Detection, and Semantic Segmentation 的更多信息

标签

产品


版本

R2017b

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by