Prediction Model of Salary Dynamic Fluctuation Trends Incorporating Multivariate Time Series

Abstract

With the continuous development of human resource management, salary management has emerged as a crucial area of enterprise focus. To better predict the dynamic fluctuation trends of salaries, the establishment of a rational prediction model is of paramount importance. To this end, a prediction model for salary dynamic fluctuation trends is constructed based on multivariate time series and the BP neural network model. Data spanning from 2006 to 2015 are selected to forecast salary fluctuations during the period from 2021 to 2025, and the prediction results are comparatively analyzed with actual index data. The research findings indicate that residents' salary income approximately synchronizes with national economic growth levels and aligns with real-world scenarios. However, significant salary fluctuations are observed, which can be attributed to adjustments in national fiscal policies and the enhanced development of the national economy, leading to an upward trend in employee salaries.