Machine learning and data analysis
Vladimir Mityushev, R. A. Kycia, Wojciech Nawalaniec, Natalia Ryłko · 2024
This section introduces machine learning and data analysis, which is currently one of the fast-paced research fields. The methods presented here are different from those presented before. They are summarized in a data-driven paradigm, where we do not create algorithms (step-by-step instructions) for producing outputs from inputs. Instead, we present some specifically prepared data to the program, which is responsible for finding patterns and generalizing them to other similar inputs. The chapter covers the description of supervised, unsupervised learning, and data preparation for machine learning. A simple example of classification problem and statistics are presented. Reading, cleaning and scaling data; dimensionality reduction by PCA; Selected models of supervised and unsupervised learning; regression and neural networks are discussed. The chapter concludes with an appendix being basic introduction to Python programming language.