Why Training Set accuracy decrease dramatically after stopping the trainNetwork?
2 visualizaciones (últimos 30 días)
Mostrar comentarios más antiguos
After stopping manually trainNetworktrainNetwork, the validation error dropped dramatically:
I tested the Training Set accuracy, and got also about 60%:
predY = classify(net,xTrain);
Any ideas what I'am doing wrong?
4 comentarios
Don Mathis
el 23 de En. de 2019
What is your network architecture? Does it contain dropoutLayers and later BatchNormlization layers?
Sergy Stepura
el 1 de Feb. de 2019
Editada: Sergy Stepura
el 4 de Feb. de 2019
Respuestas (1)
Don Mathis
el 8 de Feb. de 2019
Maybe your minibatch size is too small. The accuracy drop may be due to batchnormalization layers getting finalized, during which time the mean and variance of the incoming activations of each batchnorm layer are computed using the whole training set. If those full-batch statistics don't match the minibatch statistics very well, the finalized batchnorm layers will not be performing a very good normalization.
3 comentarios
Don Mathis
el 11 de Feb. de 2019
You could try increasing the batch size iteratively to see whether that fixes the problem. I would try exponentially increasing: 1000, 2000, 4000, 8000, etc. Or you can just try the largest amount that will fit in your GPU memory right away.
Don Mathis
el 11 de Feb. de 2019
Also: Why does your plot show "Iterations per epoch: 1"? Were you using miniBatchSize=30000 in that run?
What are you passing to trainingOptions()?
Ver también
Categorías
Más información sobre Image Data Workflows en Help Center y File Exchange.
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!