Evaluatıon Of The Effect Of Train-Test Split Ratıo For The Prediction Of Daily Pan Evaporation Of Çorum Province

This study aims to estimate the evaporation data in the Çorum station. For this purpose, the effect of the training-test ratio on machine learning model performance was evaluated using daily evaporation data from the Çankırı meteorology station. Moreover, Random Forest and GBM algorithms were used to model pan evaporation. The study’s second aim is to investigate the training test rate’s effect on pan evaporation prediction. For this purpose, the training test rates were used as 90-10%, 80-20%, 70- 30%, 60-40%, and 50-50%.