Improved Accuracy for prediction of leaf wetness using Logistic Regression algorithm compared with Decision Tree algorithm

Abstract

The main aim of this research work is to compare the accuracy percentage of leaf wetness predicted by the Novel Logistic Regression algorithm to that predicted by the Decision Tree method using meteorological data. The accuracy of leaf wetness prediction was evaluated using Novel Logistic Regression and Decision Tree with a sample size of 20 at different times. Novel Logistic Regression has a significantly better accuracy percentage (91.89%) compared to Decision Tree accuracy (80.24%). Between Novel Logistic Regression and Decision Tree, The statistical significance difference p=0.020 (p<0.05) independent sample T-test value state that the results in the research are significant. The Decision Tree method fared much worse than Novel Logistic Regression.