Comparitive Analysis of Support Vector Machine and Convolutional Neural Network in Image Classification

This research is aimed at comparing the effectiveness of two AI models in the classification of images of cats and dogs. Based on previous research in this field, two distinct models presented the best results, Support Vector Machine from traditional machine learning methods and Convolutional Neural Networks from deep learning methods. A dataset of 20,000 and a subset of 4000 images were used to train and test the models. The CNN model exhibited better results, with an overall classification accuracy of 89% and 76%, respectively. The SVM model performed poorly, with an accuracy of only 61%. This paper discusses the differences and efficacy of each model classifying pets.