EVALUATION OF PROFESSION PREDICTIONS FOR TODAY AND THE FUTURE WITH MACHINE LEARNING METHODS
In the analysis of the present and future of the professions, the fact that the data subject to the analysis is in text form reveals text data mining as an important research area. While popular occupations in the labor market can be determined by text mining instead of employer surveys, machine learning algorithms are used for predictions about the future of the professions. Machine learning models consist of two different algorithm models: supervised and unsupervised learning. In this study, supervised learning algorithms and clustering methods are applied to (Labor Market Research – LMR) report and occupations data sources and it is tried to make inferences about the relationships and properties of the data. Popular professions have been predicted with an accuracy rate between ≅0.81 and ≅0.93 with various machine learning algorithms. Perceptron and stochastic gradient descent algorithms have demonstrated superiority over other algorithms thanks to their intelligence functions. In the analysis for the professions of the future, results with an accuracy of nearly 1 were produced, despite the limitation of insufficient Turkish data sources.