The Classification of Mushroom Types Using Naïve Bayes and Principal Component Analysis

Abstract

Indonesia is one of tropical countries with high humidity which makes it possible for various plants and microorganisms to grow properly. One of the microorganisms that shall grow well in Indonesia is be considered as fungi or mushrooms. They have several types including poisonous and edible mushrooms that shall be consumed by human beings. The purpose of this research is to make it easier to classify between the types of poisonous mushrooms and edible mushrooms which can be consumed by using the Naïve Bayes algorithm to get the accurate classification results. In this research, the Naive Bayes algorithm is used to classify the types of mushrooms by utilizing the Principal Component Analyst technique which serves to reduce the number of features applied in the dataset. The data collection technique used in the research is by documenting the official website of the UCI Machine Learning Repository whereas the Mushrooms dataset consists of 22 features and 1 class are applied. After classifying using Naïve Bayes with Principal Component Analyst, then the researcher is evaluating using the 10-Fold Cross Validation technique whereas the results obtained are pc = 10 and the classification result is be considered as 84%.