Návrh a implementace řešení úlohy strojového učení v cloudu

This thesis deals with the machine learning task in the cloud. The aim of the thesis is to design and implement a solution in the cloud to estimate the probability of voluntary or involuntary dropout. The first part of the thesis is devoted to understanding the problem of academic failure – the main theoretical approaches, causes, and consequences. The next part focuses on the selection of a suitable machine learning service on the Microsoft Azure platform. From the wide range of provided services described in this section, the Azure Machine Learning service was chosen for its versatility. Furthermore, the thesis discusses predictive data analytics, machine learning, and their possible applications. Subsequently, the thesis focuses on the design and implementation of a solution to the task of estimating the probability of dropping out using Microsoft Azure platform services. The implementation of the solution was carried out in accordance with the CRISP-DM methodology.