High Performance Computing Center North
This two day course gives a practical introduction to deep learning, convolutional and recurrent neural networks, GPU computing, and tools to train and apply deep neural networks for natural language processing, images, and other applications.
The course consists of lectures and hands-on exercises. Keras and TensorFlow will be used in the exercise sessions. CSC's Notebooks environment will be used on the first day of the course, and the HPC2N Kebnekaise cluster on the second day.
Since there may be an overlap in interest, we would also like to mention that there is a DataScience Workshop in Umeå just before this course: the 6th Swedish Workshop on Data Science.
Coffee/tea breaks will be approximately 10:30-10:45 and 14:30-14:45 each day.
Day 1: Jupyter Notebooks
9-10:30 Lecture: Introduction to deep learning
10:30-11 Exercise: Introduction to Jupyter Notebooks, Keras fundamentals
11-12 Lecture: Image data, multi-layer perception networks, convolutional neural networks
13-14 Exercise: Image classification with MLPs and CNNs
14-15 Lecture: Text data, embeddings, neural NLP, recurrent neural networks
15-16 Execise: Text sentiment classification with CNNs and RNNs
Day 2: GPU cluster
9-10 Lecture: GPUs, batch jobs, using the GPU cluster (Kebnekaise)
10-12 Exercise: Image classification
13-14 Exercise: Text categorization and labelling
14-15 Lecture: GPU utilization, multiple GPUs
15-16 Exercise: Using multiple GPUs
PREREQUISITIES: The participants are assumed to have working knowledge of Python and suitable background that includes some knowledge about data analysis, machine learning, or a related field. Previous experience in deep learning is not required, but the fundamentals of machine learning are not covered on this course. Basic knowledge of a Linux/Unix environment will be assumed, as will basic programming skills.
Coffee/tea and lunch will be provided.
Please bring a laptop for the hands-on. You will need an account at HPC2N for the second day excercises. If you do not have one, you will be contacted about creating one.
Date: 22-23 November 2018
Time: 9:00 - 16:00 both days
Venue: N410 (22 November), N280 (23 November)
Instructors: Markus Koskela, Juha Hulkkonen, CSC
Deadline for registration is 15 November 2018. The number of seats is limited, and registration may close earlier if the capacity is reached.
CLOSED. We will have a reserve list. Email email@example.com to be added to it.
Please fill in this form to sign up. Fields marked with a * are required.