Upotreba 'ARIMA' modela za promet predviđanja u proceni investicionog projekta

Abstract

U savremenim analizama investicionih projekata, najkritičnija tačka je kako proceniti dnevni promet proizvodnje, ili usluga, na bazi sistema. Da bi ostvarili predviđanje, za precizno ulaganje u određenu vrstu opreme analiziran je dnevni promet za automatsko pranje vozila u skladu sa vremenskim uslovima. Prema analizama, napravljen je ARIMA model i vremensko stanje u skladu sa 'Box-Jenkins' procedurom. Zaključak je da se dnevni promet može analitički izraziti kroz dnevne vremenske uslove. Ispravnost sistema je proverena na drugom sistemu koji je postavljen u drugom gradu Srbije. Prema upoređenim rezultatima,zaključak je da se ARIMA model za sistem dnevnog prometa, predviđen zavisnom promenljivom, generalno može koristiti kao dobar za predviđanje u analizama investicija ili za kriterijum izbora u investicionim odlukama.

In the contemporary investment project analyses, most critical point is how to estimate daily turnover of production, or service, based system. In order to make prediction, for investment in certain type of equipment more accurate, daily turnover in the system for automated car wash was observed, along with weather conditions. According to observation, ARIMA model for daily turnover and weather condition is created, according to Box-Jenkins procedure. Conclusion was made that daily turnover can be analytically expressed through daily weather conditions. Validity of observation is checked on second system that is installed in different town in Serbia. According to compared results, conclusion was made that ARIMA model of system daily turnover, predicted by dependent variable, can be generally used as good predictor in investment analyses, or selective criteria for investment decisions.