Predicting the Price of Old Mobile Phones

The world's second-hand mobile phone market is growing the fastest among all mobile phone markets. The key usermotivations revolve around cost and the sustainable aspect of mobile phones. Resale price prediction of old mobilephones is complicated, considering multiple factors including but not limited to the brand, model, specifications, age,physical condition, and dynamic changes in the market. This research looks into machine learning algorithms forestimating the price of used mobile phones. For this approach, a dataset was created from major e-commerce sites eBayand Craigslist based on brand, storage, RAM, camera specs, and age. A number of machine learning models includingLinear Regression, Decision Trees, and Neural Networks were deployed to assess the predictive and accuratecomparative performance.More sophisticated models, like Decision Trees and Neural Networks, were able to outperform the basic models andlinear frameworks because they understood more non-linear relationships between the features. Among the models,Neural Networks were the best at predicting resale prices because they can model complex price predicting patternswithin the dataset. Besides the abstract market attributes like seasonality and the releasing of higher model phones, thestudy also points toward brand, specs, and condition of the device to be the most determinants of resale price. Overall,the study offers insights toward fairer and more improved priced-driven frameworks in predictive analytics for the preowned smartphones market, to the advantage of buyers and sellers.