PENGGUNAAN APLIKASI RAPID MINER UNTUK KLASIFIKASI KUALITAS APEL DENGAN ALGORITMA DECISION TREE

Abstract

The agricultural and marketing industries are highly dependent on the quality of apples, so better methods of fruit classification are needed. Using the Decision Tree algorithm used by RapidMiner software, the study classified the quality of apples based on features such as sweetness, freshness, size, and weight. The Dataset used consists of 4001 entries with 9 attributes obtained from the Kaggle platform. To ensure the consistency and accuracy of the model, the data preprocessing stage is performed. Classification Model produces accuracy of 52.76%, with high recall for the Class "Good" but low for the class "Bad". A data imbalance was identified as a major component affecting model performance. The results show how important it is to adjust algorithm parameters and balance the data to improve the accuracy and reliability of predictive models