All topics will be covered conceptually, then illustrated in open-source Python code. Conceptual overviews assume no prior experience with neural networks or machine learning. Participants will have access to all the code and will get to experiment during the workshop and afterward.

 

Time permitting, we'll cover these major topics:

* Fully connected neural networks

* Backpropagation

* Autoencoders

* Convolutional neural networks

* Recurrent neural networks, including long short-term memory

* Regularization and dropout

* Optimizers

* Hyperparameter optimization

* Neural network visualization