Züri ML #15: Deep Learning Hands-on, and Motor Dysfunction

confirmed presentations:

• Automated Assessment of Motor Dysfunction in Multiple Sclerosis, or How to Give a Computer the Eye of a Neurologist
Jonas Dorn (https://www.linkedin.com/profile/view?id=118232487), Senior Modeler at Novartis

• Deep Learning Hands-on: Roll your own Convolution Neural Net in Python
Oliver Dürr (http://www.zhaw.ch/fileadmin/php_includes/popup/person-detail.php?kurzz=dueo), Lecturer at ZHAW

Abstracts: Deep Learning Hands-on: Roll your own Convolution Neural Net in Python
There have already been several talks about deep learning and especially convolution neural networks (CNN) in our previous meetups. In this tutorial, I will give a brief introduction to CNNs and then show you how easy it is to build your own network using the python lasagne framework. Specifically I will illustrate how to create a multiple layer perceptron and a simple CNN and show how to do on the fly data augmentation to reduce overfitting.

Please bring your own laptop with python and lasagne installed! The iPython notebooks I will use are available at https://github.com/oduerr/dl_tutorial/tree/master/lasagne . For an installation instruction for lasagne, see e.g. http://lasagne.readthedocs.org/en/latest/user/installation.html#install-from-pypi