Importance of Feature Extraction in Naïve Bayes Applied on Animal Classification

Abstract

This study explores the use of the Naive Bayes classification algorithm for animal classification, particularly focusing on its computational efficiency and simplicity. The research evaluates the algorithm's performance using a dataset of 15,000 animal images across five species (cat, dog, elephant, horse, lion). The study investigates the impact of feature extraction techniques, such as Histogram of Oriented Gradients (HOG), on the classifier's accuracy. Results indicate that the Naive Bayes classifier, coupled with HOG feature extraction, performs well in terms of precision, recall, and F1-score. The classifier's simplicity and computational efficiency make it well-suited for large datasets, contributing to its scalability and usefulness in biological research. Moreover, the research emphasizes that combining Naive Bayes with other classifiers like Support Vector Machines (SVM) could mitigate some of its limitations, such as bias toward majority classes and feature independence assumptions. Overall, this study highlights the potential of Naive Bayes as a robust and efficient tool for animal classification and offers insights into its broader applicability in handling large biological datasets.