In this project, I began with data preparation and cleaning, which involved loading the necessary packages and dataset, and performing basic data cleaning steps such as removing missing values and irrelevant columns to ensure data quality. Then, I used LASSO regression for feature selection, identifying the most important features for subsequent linear regression model construction. During the modeling process, I applied feature transformation to address heteroscedasticity and handled outliers to improve model accuracy. Finally, I compared the performance of multiple models using metrics like adjusted R-squared, AIC, RMSE, and RMSE to evaluate the model's validity.