Registration is now closed!
This is a two day beginner course on machine learning for high energy physics given by Jun.-Prof. Dr.Gregor Kasieczka (Universität Hamburg), and will cover topics like
- Basics of neural networks and training
- Convolutional Networks
- Recurrent architectures
- Combining Physics and Deep Learning
- Systematic Uncertainties
- Learning from Data
- Generative Networks
- Understanding network decisions
- Hands-on exercises
The course is free of charge and consists of three lecture blocks open for a larger audience, as well as a hands-on session.
You will learn how to use tools like Keras and TensorFlow, and the hands-on session includes designing your own deep neural network (both simple as well as more complex architectures) and training it on a cloud GPU.
This course is targeted for PhD students and Postdocs, but others will be allowed to participate if there are available places. Privilege is given to physics students from UZH and ETH.
The lectures and hands-on session will take place at the UZH Irchel Campus.
On both evenings we will have invited talks given by experts in the field:
Deep learning and future challenges at HL-LHC
Jennifer Ngadiuba (CERN)
Scaling up TensorFlow on Accelerators
Marvin Ritter (Google AI)
- Thea Aarrestad (UZH)
- Ben Kilminster (UZH)
- Florencia Canelli (UZH)