Troubleshooting Python Deep Learning – additional resources

Troubleshooting Python Deep Learning – additional resources

Convolutional Neural Networks

Concatenate two CNNs:

• Our example was inspired by: https://stackoverflow.com/questions/45723596/keras-how-to-concatenate-two-cnn
• More about “merging” neural networks in Keras: https://keras.io/getting-started/functional-api-guide/#multi-input-and-multi-output-models

Splitting trained models:

• Basic splitting in Transfer Learning using pre-trained CNN model from Keras: https://medium.com/abraia/first-steps-with-transfer-learning-for-custom-image-classification-with-keras-b941601fcad5 
• Splitting in neural stye transfer: https://keras.io/examples/neural_style_transfer/
• Keras pre-trained models: https://keras.io/applications/

Resolving .fit_generator errors:

• Using .fit_generator with large image datasets: https://medium.com/difference-engine-ai/keras-a-thing-you-should-know-about-keras-if-you-plan-to-train-a-deep-learning-model-on-a-large-fdd63ce66bd2
• How to use .fit and .fit_generator: https://www.pyimagesearch.com/2018/12/24/how-to-use-keras-fit-and-fit_generator-a-hands-on-tutorial/

Model object has no attribute load_model:

• How can I save a Keras model? : https://keras.io/getting-started/faq/

High val_acc, but low accuracy in practice:

• Reducing overfitting: https://elitedatascience.com/overfitting-in-machine-learning

Error in adding a dense layer:

• The meaning of Flatten layer in CNNs: https://www.quora.com/What-is-the-meaning-of-flattening-step-in-a-convolutional-neural-network
• Using Flatten layer: https://keras.io/layers/core/#flatten

Model with multiple outputs errors:

• Complete examples of multi output network arch for image classification, Keras: Multiple outputs and multiple losses: https://www.pyimagesearch.com/2018/06/04/keras-multiple-outputs-and-multiple-losses/
• Multi input, multi output models: https://keras.io/getting-started/functional-api-guide/#multi-input-and-multi-output-models

Model that uses Dropout is still overfitting:

• Reducing overfitting: https://elitedatascience.com/overfitting-in-machine-learning
• What’s dropout and how to use it: https://machinelearningmastery.com/dropout-for-regularizing-deep-neural-networks/

Recurrent Neural Networks

Long-Short Term Memory Networks