Pre-course write-up

The notebooks and data used throughout the course are available here:

Our excellent course lecturers Igor Barros Barbosa and Aleksander Rognhaugen have also prepared possible solutions to the notebook exercises. These can be found as HTML files in the attached zip-file.

A description on how to set up a cloud computing machine with the necessary software as used during the course will be made available in due time (participants will be notified).

As some people are more familiar with R than Python, it is worth mentioning that the Keras library that was used in the course now also has an R interface. Actually there seems to be two packages: keras: and kerasR:

Note that they are both fairly new, so there might very well be some bugs at the time being.

COURSE Information

I hope you are all ready for the "Introductory course to deep learning", before the Norwegian Statistical Meeting in Fredrikstad. Below you will find some basic information about the course.

Course lecturers: Igor Barros Barbosa and Aleksander Rognhaugen

Course language: English

Where: The main conference room in Quality Hotel Fredrikstad.

When: Monday June 12th . 13.30 – ca. 18.00 and Tuesday June 13th 09.00 – 12.30.

Content: The course is an introductory course. The basic concepts of deep learning will be introduced and explained both in theory and through applications to linear regression and image classification.

Practicals: An important part of the course are the practicals where you will be building and training a complete (deep) neural network. These practicals will be carried out by connecting your computer to a cloud based system of virtual machines with GPUs. All you need to bring is a laptop with a modern web browser. The programming language will be Python, and we will work with Tensorflow through the user friendly package Keras. No prior knowledge of these systems is necessary as you will not need to write code from scratch. All code and setup will be made available on Github. It is optional whether you pair-up with a colleague or work alone on the practicals.

Course start: Monday 13.30, but please meet up 15 minutes early to check that you are able to get internet access and connect to the virtual machine. Login details will be provided by email or when you meet up before the course.

Lunch: Note that lunch at 12.30 on Monday (before the course start) is included as part of the accommodation/course package unless you have other agreements.

Other information: See the web page for the meeting