Oscillation in neural network

My neural network is consist of following layer for dataset of 8336 * 364

 featureInputLayer(numFeatures,'Normalization', 'zscore')
    fullyConnectedLayer(50)
    batchNormalizationLayer
    reluLayer
    fullyConnectedLayer(numClasses)
    softmaxLayer
    classificationLayer
   

During training process there is oscillation in accuracy of the network. I want to Know How can I improve my network.

My validation accuracy is 80,64 and test accuracy is 83%.

I used regularization and dropout

Answer 1

[Srivardhan Gadila]

You can refer to the documentation of Deep Learning Tips and Tricks & resources available under Deep Learning Tuning and Visualization.