Infopaket

Software and computer complexes for high-energy physics

ВБ 3.3

Вибіркова дисципліна для ОП

Second

2018/2019

3 Semester

6

1. Know the principles of neural networks, basic architectures, principles of data preparation, training and evaluation of the result 2. Practical skills in building and training neural networks.

Full-time form

- Successful learning of basic physics courses: "Mathematical Analysis", "Linear Algebra" "Mechanics", "Molecular Physics", "Electricity", "Optics". - Be able to solve problems in the main physics and mathematics courses. - Have the profound skills of working on a computer to search for information on the Internet. - The student must know the Python programming language and its main libraries

The course "Software and Computer Complexes for High Energy Physics" will significantly improve professional training of students of the Nuclear Physics Department, which is due to the fact that the students will be: - know the principles of neural networks construction and algorithms and methods of their training. - Know how to prepare data for training and verification of neural networks; - Have knowledge of the architectural principles of neural networks - To perform architectural arrangements for training and prediction modes in neural networks.

1. Deep Learning. deeplearningbook.org 2. Practical Deep Learning for Coders. https://www.fast.ai/ 3. CS231n: Convolutional Neural Networks for Visual Recognition. cs231n.stanford.edu

Lecture demonstration, practical exercises, self-study; consultations

- semester assessment: 1. the assessment during the lectures (maximum 50 points). 2. Practical exercises (maximum 50 points). - Summative assessment in the form of an examination

Ukrainian