Enhancing Cascading Style Sheets Efficiency and Performance Through AI-Based Code Optimization

This paper proposes an AI-based method of CSS improvement that would result in efficiency and performance enhancement in real-world applications based on CSS. The core of this study revolves around using machine learning strategies like SVM, Naive Bayes, Ensemble Methods, and Logistic Regression for the task of CSS code classification and optimization to bring out a better quality user interface. The dataset applied was three-category labelled data, with the application of preprocessing techniques, including feature extraction and normalization, for the accurate training of the model. Each model's performance was expressed based on critical metrics, which include ROC-AVC, Fl score, Recall, Accuracy Accuracy, and Precision. Based on the models tested, the AI-based proposed model exhibited better predictability, with the highest AccuracyAccuracy of 97% and the highest Fl score of 97% compared to the ones concerned. The experiment's results show that incorporating AI-based optimization significantly improves the quality of the CSS code, bringing better classification performance and error detection- a case wherein the proposed model outperformed all other traditional methods. The above findings give insight into the possibility of AI-driven solutions in optimizing web performance and can even improve the quality of the CSS code in various applications.