Research on the Prediction Models Based on Mobiles Dataset

Haoning Li, Zhexi Wang, Haoxuan Yang

2025

Abstract

This study utilizes multivariate linear regression (MLR) and random forest (RF) models to predict smartphone market trends, analyzing a dataset of 930 models with specifications like Random Access Memory (RAM), cameras, battery capacity, and regional prices. The goal is to decode how hardware features and pricing strategies influence market dynamics, offering data-driven insights for industry stakeholders. MLR was applied to explore linear relationships between features and China launch prices, while RF modeled nonlinear patterns. The dataset was split into 80% training and 20% test subsets, evaluated via R² and RMSE. Feature importance in RF highlighted key predictors. Findings show MLR identifies RAM, mobile weight, and screen size as significant linear predictors but with limited explanatory power. RF outperforms, demonstrating stronger training fit and generalization, with front camera and RAM as top drivers. Complex interactions emerge, such as positive effects of screen size/weight and negative impacts of battery capacity on prices. Conclusively, RF excels in capturing non-linear trends, while MLR provides foundational linear insights. Both models underscore RAM, camera specs, and screen size as critical pricing determinants. The results guide manufacturers in feature prioritization and pricing strategies, with R² and RMSE validating model robustness for market trend analysis. These insights enhance data-informed decision-making in the dynamic smartphone industry.

Paper Citation

in Harvard Style

Li H., Wang Z. and Yang H. (2025). Research on the Prediction Models Based on Mobiles Dataset. In Proceedings of the 2nd International Conference on Innovations in Applied Mathematics, Physics, and Astronomy - Volume 1: IAMPA; ISBN 978-989-758-774-0, SciTePress, pages 348-354. DOI: 10.5220/0013825600004708

in Bibtex Style

@conference{iampa25,
author={Haoning Li and Zhexi Wang and Haoxuan Yang},
title={Research on the Prediction Models Based on Mobiles Dataset},
booktitle={Proceedings of the 2nd International Conference on Innovations in Applied Mathematics, Physics, and Astronomy - Volume 1: IAMPA},
year={2025},
pages={348-354},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013825600004708},
isbn={978-989-758-774-0},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 2nd International Conference on Innovations in Applied Mathematics, Physics, and Astronomy - Volume 1: IAMPA
TI - Research on the Prediction Models Based on Mobiles Dataset
SN - 978-989-758-774-0
AU - Li H.
AU - Wang Z.
AU - Yang H.
PY - 2025
SP - 348
EP - 354
DO - 10.5220/0013825600004708
PB - SciTePress