Application-Oriented Deep Learning in Physics
Syllabus, Master's level, 1FA368
- Code
- 1FA368
- Education cycle
- Second cycle
- Main field(s) of study and in-depth level
- Physics A1N
- Grading system
- Fail (U), Pass (3), Pass with credit (4), Pass with distinction (5)
- Finalised by
- The Faculty Board of Science and Technology, 25 March 2021
- Responsible department
- Department of Physics and Astronomy
Entry requirements
120 credits in science/engineering with Computer Programming I, Quantum Physics or Solid State Physics I/Statistical Mechanics. Proficiency in English equivalent to the Swedish upper secondary course English 6.
Learning outcomes
On completion of the course the student shall be able to:
- summarize the concepts of deep learning
- apply deep learning for typical physics problems
- design and optimize network architectures for different problems in physics and engineering
Content
Fundamentals of Deep Learning. Generalization, Regularization and Validation. Optimization and Hyperparameter Tuning; Convolutional Neutral Networks; Classification and Regression Tasks; Visualization & Advanced Computer Vision Methods; Autoencoders; Generative models, variational autoencoders, generative adversarial networks. Applications in different fields within physics, for example the design a trigger algorithm to measure cosmic neutrinos with a detector in Antarctica, and analyze data of the world's largest particle accelerator LHC. Analysis of big data sets in experimental physics. Astroparticle physics, elementary particle physics, image recognition, de-noising of data. Practical skills of using the TensorFlow framework via the high-level Keras python interface.
Instruction
Lectures and exercise classes.
Assessment
Hand-in problems.