PRICEVISION: DEEP LEARNING-DRIVEN PREDICTION OF USED VEHICLE RESALE VALUES

Abstract

The used car market has become one of the fastest-growing sectors in the global automobile industry, with consumers demanding reliable price estimates before purchase or sale. Traditional valuation methods rely heavily on manual appraisal and subjective judgment, leading to inconsistent results. This paper introduces PriceVision, a deep learning-driven framework designed to predict used vehicle resale values with high precision. By leveraging data such as vehicle make, model, age, mileage, fuel type, transmission, and market trends, PriceVision utilizes advanced neural network architectures to capture non-linear relationships within large datasets. The system employs feature normalization, one-hot encoding, and regression-based learning to achieve optimized predictions. Experimental results demonstrate that PriceVision outperforms conventional machine learning methods in terms of accuracy, reliability, and adaptability to market fluctuations. This study emphasizes how deep learning can transform the automotive resale sector by offering intelligent, data-driven decision support for both buyers and sellers.