USING ARTIFICIAL NEURAL NETWORK TO PREDICT PROFESSIONAL MOVEMENTS OF GRADUATES / UPORABA UMJETNIH NEURONSKIH MREŽA U PREDVIĐANJU PROFESIONALNIH KRETANJA MATURANATA

Abstract

ABSTRACT

 Aim of this paper is to examine the possibility of predicting professional choice of secondary school graduates by using artificial neural network for the needs of enrolment policy planning at the Faculty of Mechanical Engineering in Belgrade. The problem of professional choices prediction for graduates has been carried out on the sample of 119 graduates from two Belgrade vocational schools. Factors influencing professional choice of secondary school students grouped in twelve input variables proved to be as suitable for predicting professional movements of graduates. Results of this research represent the base for the further researches aiming to improve the Faculty enrolment policy.

Keywords: predictive techniques, faculty enrolment policy, prediction of professional choice

-----

SAŽETAK

Svrha ovog rada je istražiti mogućnosti predviđanja izbora profesije maturanata uporabom umjetnih neuronskih mreža zbog potrebe planiranja upisne politike Strojarskog fakulteta u Beogradu. Istraživanje predviđanja odabira profesije provedeno je na uzorku od 119 maturanata dvije beogradske strukovne srednje škole. Svrstavanje čimbenika koji utječu na izbor profesije kod srednjoškolaca u dvanaest ulaznih varijabli pokazalo se odgovarajućim za predviđanje profesionalnih kretanja maturanata. Rezultati ovog istraživanja predstavljaju osnovu za daljnja istraživanja usmjerena poboljšanju upisne politike Fakulteta.

Ključne riječi: tehnike predviđanja, upisna politika fakulteta, predviđanje izbora profesije