Laptop Price Estimation Using Data-Driven Predictive Analytics

In today’s fast-paced world, selecting a laptop that provides best performance and durability under budget is a challenge for consumers. This paper is based on predicting prices of laptops based on some inputs or specifications inputted by the user. It uses a cleaned dataset of laptops — 1,300 records — filtering out duplicates and missing variables. It has been tested for different models like XGBoost, Random Forest, Stacking model, Linear Regression, SVM, Ridge Regression, Gradient Boost, KNN, Decision tree, etc. These models investigated the cost implications of laptop specifications like brand, type, RAM, weight, touchscreen, screen size, screen resolution, CPU, HDD, SSD, GPU and operating system. Performance was being evaluated using R2 score and Mean Absolute Error (MAE). According to the results, Random Forest outperformed the others, obtaining the highest R2 score and the lowest MAE. This study assists consumers in making informed selections, and allows retailers to combine data-driven pricing strategies.