The modern automobile industry is constantly evolving, and consumer preferences are significantly influenced by design aspects. This paper aims to analyze the impact of car design on sales using Convolutional Neural Network (CNN). For the study, images of the top 100 cars with the highest sales in 2022 were collected to extract powerful visual features. The trained CNN model identifies the correlation between specific elements of car design and sales, and constructs a predictive model.

Additionally, the model is utilized to collect and analyze price and fuel efficiency data of the top 100 cars. This provides objective numerical data for comparing designs and evaluating the predictive results of the model reliably. The experimental results highlight the importance of design elements and are expected to offer insights into marketing strategies and product development in the automobile industry. Moreover, to verify the model's generalization capability, the predictive performance is evaluated using arbitrary car data, not from the top 100 cars with high sales in 2022. This is crucial for confirming the model's ability to categorize cars and predict sales within those categories.

Keywords

Car Design, Sales Prediction, CNN, Visual Features, and Consumer Preferences,