What type of study is this?

September 10, 2025

Research on e-commerce retail demand forecasting based on multi-model fusion

Puntos clave

The model achieves robust demand forecasting for e-commerce, enhancing inventory management.
Results show high prediction precision and stability with the fused ARIMA and GM models.
Data preprocessing with outlier elimination ensures more accurate insights from e-commerce statistics.
K-Means clustering effectively extracts key features for better forecasting outcomes in retail.

Resumen

With the rapid refinement of the e-commerce retail sector, accurately predicting the demand for commodities has become the key for enterprises to optimize inventory management and improve supply chain efficiency. From current research, we propose an isodimensional new interest recurring GM model, ARIMA model, and K-Means clustering multi-model fusion method for e-commerce retail demand forecasting. Firstly, the etailing statistics was cleaned and preprocessed, and the 3-sigma criterion was used to identify and eliminate outliers, and secondly, the isodimensional new information compensatory GM model and the ARIMA model were combined for preliminary prediction, and the key features were extracted by K-means clustering analysis. The findings indicate that the model can accurately and robustly predict the demand in e - commerce retail. It shows high prediction precision and stability. The model developed in this study is able to offer technical assistance for the demand forecasting within the e - commerce retail domain.

Preguntar a la IA

Me gusta

Guardar