Diamond Price Prediction Based on Regression Model: Comparison and Analysis of Different Algorithms

Abstract

Diamonds, as a valuable commodity, have investment value. Through price prediction, investors can determine the trend of diamond prices and decide when to buy or sell diamonds to achieve maximum investment returns. For some consumers who plan to purchase diamonds as collectibles or investments, price forecasting can help them determine whether the purchased diamonds have appreciation potential and make more informed consumption decisions. The government or relevant industry associations can formulate corresponding industrial policies based on the trend and prediction of diamond prices, regulate market order, and promote the healthy development of the diamond industry. In order to compare and analyze the performance of several regression models in diamond price prediction, and which one used, this study preprocessed multiple features, including standardization and feature selection, on the diamond dataset on Kaggle. Subsequently, this paper compares the performance of linear regression, random forest, XGBoost, and Deep Neural Network (DNN) in price prediction tasks. This study indicates that the random forest model has superior performance in diamond price prediction and can be used for actual market pricing.