<![CDATA[Fungi are heterotrophic organisms that fulfill their needs by utilizing organic matter from other organisms. Mushrooms consist of 3 groups, namely edible mushrooms, poisonous mushrooms, and mushrooms that are not known to be edible or poisonous. The purpose of this research is to find the best algorithm to classify mushroom types appropriately. Mushroom classification can be done through the application of a single classifier including the Gaussian Naive Bayes algorithm, C4.5 and Classification and Regression Trees (CART), which is then compared using an ensemble classifier. Single classifiers have limited performance and are not adaptive to data changes, while the ensemble algorithm with Weighted Soft Voting is able to cover these shortcomings by increasing accuracy and robustness by combining predictions from several models that are weighted based on their respective performance. The Gaussian Naive Bayes, CART and C4.5 algorithms with percentage split testing techniques obtained accuracy results of 63.55%, 97.67% and 97.77% respectively. The results of research using the Ensemble method can increase the accuracy value by combining the three algorithm weights with an accuracy result of 98%. In this case, the Ensemble algorithm obtained the best performance in mushroom classification.]]>
Copyrights © 2025
