Модифiкованi баєсiвськi класифiкатори для текстових повiдомлень
Д. Фітель · Electronic Kyiv-Mohyla Academy Institutional Repository (National University of Kyiv-Mohyla Academy) · 2017
The traditional methods of statistical classification of text spam are analyzed. For a naive Bayesian classifier, a number of modifications are proposed. Based on these modifications, an alternative algorithm is described and its time and space complexity are characterized.