CLUSTERING. DECREASE OF DIMENSION OBJECT DESCRIPTION IN MASTER EDUCATION AND VISUALIZATION OF DATA. T-SNE METHOD
Abstract
In this article, the question of clustering, reducing the size of the characteristics of the sample of input objects of analysis and data visualization using the t-SNE algorithm. A variant of possible improvement of this algorithm is proposed. For testing, Python language and the popular scikit-learn library are used.
Key words: machine learning, clustering, diminishing of the dimension of the sample of traits, data visualization, t-SNE algorithm.
Fig .: 1 .. Bible: 5.