Machine Learning

computer science
mathematics
physics

The student should know the basic concepts of algebra and mathematical analysis.

The student should be able to program in python.

Classroom

Lecture and exercises introduce the students to the subject and methodology of machine learning and modeling of artificial neural networks, and to solving practical problems with these tools. The lecture is intended for third-year students of Neuroinformatics.

Program:

1. Introduction: linear regression and least squares method

2. Classification and logistic regression

3. Generative algorithms

4. Support vector machines

5. Introduction to neural networks, linear neural networks

6. Rosenblatt's Perceptron

7. Differentiable non-linearities and back error propagation

8. Deep neural networks

9. Unattended learning

10. Learning with reinforcement

Issues discussed theoretically during the lecture will be illustrated in practice with practical examples in the python language.

Knowledge:

1. The student knows the basic concepts related to machine learning and artificial neural networks (KW01);

2. has knowledge in the field of higher mathematics and information technology necessary to solve physical problems of medium complexity using machine learning methods (KW02).

Skills:

1. The student is able to apply a machine learning approach or an artificial neural network to a practical problem (KU01);

2. can perform simple experiments, observations, numerical calculations and computer simulations using standard software packages and critically analyze the results of measurements, observations, and calculations along with the assessment of the accuracy of results (KU03).

attitudes:

1. The student appreciates the importance of machine learning methods in modern methods of data analysis (K_K06);

2. The student appreciates his own work in deepening knowledge and skills in machine learning (K_K01);

3. The student is able to properly define the priorities for the implementation of specific tasks and projects of a diverse nature (K_K03).

Expected student workload:

Participation in classes: 60 hours

Preparation for classes and solving homework assignments: 20 h

Preparation for the 10-hour exam

Preparation of the final project 20h

The grade is the average of the test result on theoretical issues and the practical tests.

Attendance at the lecture is not obligatory.

Two absences from the exercises are allowed.