A gentle introduction to Deep Learning

Get a basic theoretical foundation regarding deep learning and some hands-on experience running feed forward neural network models, with examples and short programming assignments, i.e. fill in the blanks code. Answers are provided as well.

Goal

Computer vision and natural language processing topics are not discussed, but you are welcome to join the deep learning for computer vision and natural language processing course.

Content

machine learning essentials
neural networks
activation functions
training networks
back-propagation, loss functions, categorical cross-entropy, gradient descent, vanishing gradients, weight initialisation, batch normalisation

Prerequisits

Python programming. You’re also welcome to join us if you have coding experience in other languages, in which case you can co-op with a peer or go through the programming answers rather than do them as exercises.

You will work in the cloud and don't need software installed other than Google Chrome - or other browsers at your own risk (make sure to have a backup chrome installed before the course). So bring your own laptop or use one of the onsite computers. We will send you two small assignments in advance, to avoid teething issues in your setup during the day.

Enrollment and course fee

Email datascience rug.nl to enroll

You can enroll/unenroll until Monday the 15th of April. The participants will be mailed the course material Tuesday the 16th.

For more info, you can mail Herbert Kruitbosch (h.t.kruitbosch@rug.nl).