Infopaket

Image Recognition and Cluster Analysis

ВК2.2

Вибіркова дисципліна для ОП

First

2021/2022

5 Semester

5

Being able to use methods of numerical intelligence, machine learning, fuzzy and neuron network data processing, genetic and evolutionary programming for solving problems of recognition, prediction, classification and identification of control objects. Know how to employ methods and algorithms of numerical intelligence and intellectual data analysis in problems of classification, prediction, cluster analysis, search for associative rules with use of programming tools of support of multidimensional data analysis on base of DataMining, TextMining, WebMining technologies. Possess knowledge and apply data and artificial intelligence intellectual analysis methods, which include methods of computer linguistics, deep learning, neuron network technologies and computer vision methods.

Full-time form

1. Possess knowledge of algorithm theory fundamentals, basis of operations with neuron networks. 2. Being able to design and program data processing algorithms, create and research neuron networks. 3. Possess skills as to work with mathematical packages, fundamentals of programming on the languages Python and/or R.

This subject pays main attention to considerations of the main notions and methodology of image recognition, principles of creation and technologies of image recognition system design, trends and perspectives of image recognition system development, gaining knowledge of the models and methods of image recognition problem solution. This subject deals with the following methods and tasks: features generation and selection methods, deterministic methods, methods of image recognition with use of cluster analysis, statistical methods and classifiers on the base of Bayes decision making theory. The subject offers linguistic, geometrical, logical methods as well as the method of potential functions. Attention is paid to the ability to use image recognition systems for applied tasks solution in different subject areas, design image recognition systems.

2. Charu C. Aggarwal, Chandan K. Reddy Data Clustering): Chapman and Hall/CRC, 2013. – 704 p. 3. Machine Vision and Image Recognition. Ed. Jovan Pehcevski, Oakville Canada, Arcler Press, 2020 – 321 p. 5. Advance Concepts of Image Processing and Pattern Recognition: Effective Solution for Global Challenges, Eds. Narendra Kumar, Celia Shahnaz, Krishna Kumar, Mazin Abed, Mohammed Ram, Shringar Raw – Singapore: Springer Nature, 2022. – 225 p.

Lectures, laboratory classes, individual work

Student assessment is performed over the semester in all types of activities, including study of the course theoretical material, completion of the laboratory works and independent work. After completion of the module 2 topics, the students are given written control paper, which includes theoretical topical questions in open form and practical tasks with calculation of sample numerical parameters and estimation of unknown parameters of the general sample. To gain positive resulting mark on the subject, students are supposed to achieve not less than 60% from the maximum available number of points. Maximum number of points that a student can obtain over the whole semester comes to 100 points.

Ukrainian