A collection of simple python scripts examining the numerous hyperparameters of deep neural networks. The tutorials are self-contained exploring one aspect at a time of how to tune deep neural networks to get better learning, generalization and prediction performance.
The tutorials stem from the eBook: Jason Brownlee, Better Deep Learning, Machine Learning Mastery, Available from https://machinelearningmastery.com/better-deep-learning/, accessed January 25th, 2019.
The contribution of this repository is the extensions of the tutorials to further explore and discuss a given topic.
The tutorials are divided into three parts:
Better Learning. Discover the techniques to improve and accelerate the process used to learn or optimize the weights of a neural network model.
Better Generalization. Discover the techniques to reduce overfitting of the training dataset and improve the generalization of models on new data.
Better Predictions. Discover the techniques to improve the performance of final models when used to make predictions on new data.
Go to code samples and discussion of the extensions
Go to code samples and discussion of the extensions
Go to code samples and discussion of the extensions
Go to code samples and discussion of the extensions
Go to code samples and discussion of the extensions
Go to code samples and discussion of the extensions
Go to code samples and discussion of the extensions
Go to code samples and discussion of the extensions
Go to code samples and discussion of the extensions
Go to code samples and discussion of the extensions
Go to code samples and discussion of the extensions
Go to code samples and discussion of the extensions
Go to code samples and discussion of the extensions
Go to code samples and discussion of the extensions
Go to code samples and discussion of the extensions
Go to code samples and discussion of the extensions
Go to code samples and discussion of the extensions
Go to code samples and discussion of the extensions
Go to code samples and discussion of the extensions
Go to code samples and discussion of the extensions
Go to code samples and discussion of the extensions