<![CDATA[Cat breed identification is often challenging due to visual similarities between breeds, yet accurate recognition is crucial for proper care. This study aims to develop an accurate cat breed classification system using a Convolutional Neural Network (CNN) algorithm with a transfer learning approach. The model was built using the MobileNetV2 architecture on a dataset consisting of 2,387 images from 12 cat breeds. The research stages included data collection, pre-processing, model construction and training, and evaluation. Evaluation results on test data showed that the developed model achieved an accuracy of 84.52%. The model demonstrated superior performance in several classes with unique visual characteristics, but still faced challenges in other classes with similar visual characteristics. These results demonstrate that the CNN method with transfer learning is highly effective and competitive for cat breed classification tasks, with room for further development to improve performance in difficult-to-distinguish classes.]]>
Copyrights © 2025
