A Multi-Model Predictive Analysis of China's Pet Industry Development: Market Trends, Global Demand, and Strategic Responses to Foreign Economic Policies
Main Article Content
Keywords
multivariate linear regression model, nonlinear regression model, polynomial regression model, support vector regression (SVR), random forest regression model, t-test
Abstract
This study aims to analyze the development trajectories of China’s pet industry and its pet food sector, forecast future market demands, and quantitatively assess the impact of foreign economic policies to formulate sustainable development strategies. By integrating multivariate linear regression, nonlinear regression, and polynomial linear regression models within a weighted framework optimized via a particle swarm optimization algorithm, we first evaluate China’s pet industry development based on pet populations and market sizes, predicting that by 2026, the number of pet cats and dogs in China will reach approximately 93.95 million and 42.66 million, respectively. To forecast global pet food demand, we apply cubic spline interpolation to supplement missing international data and construct a similar weighted model incorporating per capita GDP, projecting pet population trends across seven major markets. Focusing on China’s pet food sector, we develop a multiple linear regression model to predict production value, alongside a hybrid weighted model combining support vector regression (SVR) and random forest regression to forecast export values, yielding expected production and export values of 5,993.98 and 46.79 (in 100 million USD) by 2026, respectively. All models are validated using mean absolute percentage error (MAPE), root mean square error (RMSE), and comparisons with auto regressive integrated moving average (ARIMA) models. To quantify the influence of foreign tariff policies, we conduct correlation analysis revealing a coefficient of up to 0.87 between exchange rates and imports, then employ an SVR model to predict that China’s pet food exports would reach 44 (100 million USD) by 2028 under policy shifts. An independent samples t-test indicates a significant difference in China’s pet food export values before and after 2021. Consequently, we propose feasible strategies for the sustainable development of China’s pet food industry, including market diversification and supply chain optimization to mitigate external policy risks.
References
- [1] Sakib, M., Ahmad, S., Anwar, K. et al. Optimizing support vector regression using grey wolf optimizer for enhancing energy efficiency and building prototype architecture. Cluster Comput 28, 60 (2025). https://doi.org/10.1007/s10586-024-04802-y
- [2] Iranzad, R., Liu, X. A review of random forest-based feature selection methods for data science education and applications. Int J Data Sci Anal 20, 197–211 (2025). https://doi.org/10.1007/s41060-024-00509-w
- [3] Mamba, E. Natural Resource Rents, Manufacturing Trade, and Manufacturing Growth: Evidence from Linear and Nonlinear Regressions for ECOWAS Countries. Comp Econ Stud 67, 150–185 (2025). https://doi.org/10.1057/s41294-024-00234-0
- [4] Dong, A., & Lee, S.-K. (2024). The Study of an Improved Particle Swarm Optimization Algorithm Applied to Economic Dispatch in Microgrids. Electronics, 13(20), 4086. https://doi.org/10.3390/electronics13204086
- [5] Wekalao, J., Mandela, N., Muheki, J. et al. Design and Analysis of a Terahertz Metasurface-Based Refractive Index Sensor for Hemoglobin Detection With Behaviour Prediction Using Polynomial Regression. Plasmonics 20, 2123–2152 (2025). https://doi.org/10.1007/s11468-024-02445-8
Article Sidebar
This work is licensed under a Creative Commons Attribution 4.0 International License.