Metaverse: Design of the Car Price Prediction Model Through a Machine-learning Approach

This paper proposes a prediction model of car prices in the metaverse, using a machine-learning method. With the advent of the metaverse, virtual assets have gained significant value, and cars are no exceptions. In this context, predicting car prices in the metaverse is a great of interest for both buyers and sellers. To develop the prediction model, we use real data obtained from Kaflix company, working on the rent-a-car business. To extract the key features impacting the rental car prices, we designed various machine learning algorithms such as, random forest regression, multi-layer-perceptron, convolution/recurrent neural networks and autoregressive-moving-average model. The experimental demonstrates the rental car price could be the most significant feature to predict the rental price. Overall, the proposed model provides valuable insights into the rental car prices in the metaverse and could be used by the rental car companies for their business. One of the metaverse platforms, ZEPETO, provides a service allowing users to try to test-drive of a car, which would be the place where the model could be applied.