Визуализация многомерных данных с использованием кривых Эндрюса

Abstract

In a recent decades processing power of computer systems has had a significant surge. So now we are capable of processing that huge amount of data which has been collected by various information systems around the globe since first databases were implemented. A task of multidimensional data analysis and processing is an important area of data science, as no matter what your data describes, it is likely to have more than one parameter. Crucial part of any data analysis is its visualization. It helps researcher understand what kind of data he is working with, does it split into any classes, does it contain any outliers and so on. In this paper the use of Andrews curves in multidimensional data visualization as part of its primary analysis is described. As an example, visualization of Wheat Seeds Dataset via Andrews curves is given.