<![CDATA[In the modern era, online shopping for clothing presents the challenge of determining the correct size. This research aims to predict clothing sizes using Polynomial Regression, which can capture the non-linear relationships between body metrics and clothing sizes. The study utilizes a dataset from Kaggle comprising weight, height, age, and clothing size attributes. Through data preprocessing, including feature transformation, engineering, selection, and cleaning, the dataset is prepared for analysis. Various models are evaluated, and Polynomial Regression is identified as the most effective, achieving an R² score of 0.70755203. Hyperparameter tuning using GridSearchCV further optimizes the model, resulting in a final R² score of 72.555511% with degree 5 and alpha 1. The evaluation indicates that while the model accurately predicts sizes, it sometimes struggles with adjacent sizes, particularly in medium ranges. This research demonstrates the potential of Polynomial Regression in improving the accuracy of clothing size predictions, thereby facilitating better online shopping experiences. ]]>
Copyrights © 2025
