Development of Machine Learning Techniques with Borderline-SMOTE

Abstract

In the age of big data and advancing computing power, the dominance of deep learning has been established worldwide. Traditional image classification techniques struggle to handle vast image datasets and fail to meet the demanding standards for accuracy and speed in image classification. Convolutional Neural Networks (CNNs) have emerged as a breakthrough in overcoming these limitations, swiftly becoming the leading algorithm for image classification. Effectively harnessing CNNs for image classification has thus become a focal point of research in computer vision globally.This paper provides an overview of the research background, significance, and current status of CNN models and image classification. We delve into two prominent image classification methods based on ResNet and ShuffleNet, conducting a comprehensive
exploration of their construction methodologies and distinctive characteristics. Finally, we conduct a comparative analysis of
the performance of these two classification models.