Apple Perfection: Assessing Apple Quality with Waterwheel Plant Algorithm for Feature Selection and Logistic Regression for Classification

Abstract

This study concentrates on the evaluation of apple quality, which is a vital part of the agricultural industry. The quality of apples is examined through several factors, such as the cultivation techniques, the harvesting methods, and the post-harvest procedures. The dataset, titled "Apple Perfection," contains important characteristics such as the size, weight, sweetness, crunchiness, juiciness, ripeness, acidity, and overall quality of the apple. To make the apple quality prediction more accurate, we used different feature selection algorithms, mainly the binary Waterwheel Plant Algorithm (bWWPA), which, in fact, had the lowest average error of 0.52153, and several of the types of classification models, especially Logistic Regression, which had the highest accuracy of 0.88625. The attribute selection process found the most important attributes, which cut down the dimensionality, and hence, the model performance became better. The results of the study show that the combination of bWWPA for feature selection and logistic regression for classification can predict apple quality with high accuracy. This way of dealing with the problem gives us information that is useful for the improvement of the cultivation techniques and the post-harvest handling to the extent that we will be able to have the best quality apples. The findings of this research have a great impact on the farming industry, meaning a strong way to evaluate the quality of apples.