Subjects

Your browser does not support JavaScript, or its support is disabled. Some features may not be available.

Machine learning in physics - NFPL061

Title:	Strojové učení ve fyzice
Guaranteed by:	Department of Condensed Matter Physics (32-KFKL)
Faculty:	Faculty of Mathematics and Physics
Actual:	from 2023
Semester:	winter
E-Credits:	4
Hours per week, examination:	winter s.:2/2, C+Ex [HT]
Capacity:	unlimited
Min. number of students:	unlimited
4EU+:	no
Virtual mobility / capacity:	no
State of the course:	taught
Language:	Czech, English
Teaching methods:	full-time
Teaching methods:	full-time

Guarantor:	RNDr. Pavel Baláž, Ph.D. RNDr. Martin Žonda, Ph.D. doc. RNDr. Tomáš Novotný, Ph.D.
Classification:	Physics > Solid State Physics

Opinion survey results Examination dates WS schedule Noticeboard

Annotation -

Last update: Mgr. Kateřina Mikšová (13.05.2022)

The lecture will provide a practical introduction into basic numerical optimization techniques and machine learning methods used in classical and quantum physics as well as in other fields of science. The most important methods will be analyzed in detail during the exercises in a form of hands-on sessions and projects by using the Python libraries Scikit-learn, sktime, Tensorflow, Keras, and NetKet.

Aim of the course -

Last update: Mgr. Kateřina Mikšová (13.05.2022)

The aim of the course is to master the basic optimization techniques and methods of machine learning for use in physics and other fields of science.

Course completion requirements -

Last update: Mgr. Kateřina Mikšová (13.05.2022)

To obtain the credit, which is a condition for admission to the exam, it is necessary to collect at least 65% of points from the assignments. The questions in the exam are based on the syllabus.

Literature -

Last update: Mgr. Kateřina Mikšová (13.05.2022)

1. V. Kvasnička: Evolučné algoritmy, STU Bratislava (2000).

2. F. Chollet: Deep learning v jazyku Python. Knihovny Keras, Tensorflow. Grada (2019).

3. T. M. Mitchell: Machine learning, McGraw-Hill Science/Engineering/Math (1997).

4. A. Geron: Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow: Concepts, Tools, and Techniques to Build Intelligent Systems, O'Reilly Media (2019).

5. P. Mehta, M. Bukov, C.-H. Wang, A. G. R. Day, C. Richardson, C. K. Fisher, and D. J. Schwab, A High-Bias: Low-Variance Introduction to Machine Learning for Physicists, Physics Reports 810, 1 (2019).

6. Anna Dawid et al.: Modern applications of machine learning in quantum sciences, arXiv:2204.04198 [quant-ph] (2022).

Syllabus -

Last update: Mgr. Kateřina Mikšová (16.05.2022)

1. Crash course in Python and its libraries NumPy, SciPy, and pandas.

2. Optimization problems in physics and their solutions. Gradient methods. Stochastic optimization algorithms (hill climbing and evolutionary algorithms).

3. Basics of Monte Carlo methods. Ising and Heisenberg model. Metropolis algorithm, heat bath algorithm. Ergodic theorem, detailed balance condition. Simulated annealing.

4. Basic methods in machine learning. Linear regression, logistic regression, support vector machines, decision trees, random forests.

5. Feed forward neural networks. Supervised learning. Backpropagation algorithm.

6. Unsupervised learning of neural networks. Hopfield neural networks. Boltzmann machines. Restricted Boltzmann machines. Autoencoders. Automatic phase classification.

7. Deep learning. Convolutional neural networks. Neural network regularization. Image recognition.

8. Analysis and forecasting of time sequences. Arima model. Recurrent neural networks. LSTM and GRU memory cells.

9. Application of neural networks in quantum physics. Neural network quantum states and quantum state tomography.

10. Neuromorfic computing. Basic concepts and current state of the research in the field.