Solving A ML Problem Using The Grossone

Abstract

Machine learning (ML) has grown at a remarkable rate, becoming one of the most popular research directions. It is widely applied in various fields, such as machine translation, speech recognition, image recognition, recommendation system, etc. Optimization problems lie at the heart of most machine learning approaches. So, the essence of most ML algorithms is to build an optimization model and learn the parameters in the objective function from the given data. A series of effective optimization methods were put forward, in order to promote the development of ML. They have improved the performance and efficiency of ML methods. The aim of this paper is to show that, among many other fields, the grossone may be used successfully in the ML. The grossone, the infinite unit of measure, has been proposed by Professor Y. Sergeyev in a number of noticeable works, as the number of elements of the set, N, of natural numbers. It is expressed by the numeral . This new computational methodology would allow one to work with infinite and infinitesimal quantities in the ―same way‖ as that working with finite numbers More details about it are given in Section 4. We analyze the SVM from the viewpoint of mathematical programming, solving a numerical example using the grossone. The Iris dataset was chosen for the implementation of the support vector method. This is a wellknown set of data used in the area of ML.