Accurate prediction of employment rate of graduated students can greatly help education authorities to make informed decisions as well as for universities to adjust their teaching plans. Unfortunately, prediction of the employment rate of graduated students is still a difficult problem because the historical employment rate data exhibits fluctuations. In this paper, a new method is proposed for employment rate prediction using fractional gray GM(1,1sin) model, which aims to alleviate the effect of data fluctuation on prediction accuracy and increase the contribution of new data in the prediction procedure. To further decrease prediction error, a Fourier series is adopted to model the residual series. The proposed model, called Fourier Fractional GM(1,1sin) Model (FFGMsin), is used to predict the employment rate of graduated students of Yanshan University from 2010 to 2019. Results show that the proposed method can obtain more accurate prediction results than GM(1,1) model and its variants. © 2024 Institute of Electrical Engineers of Japan. Published by Wiley Periodicals LLC.

中文翻译：

---

将傅里叶分数 GM(1,1|sin) 模型与大鼠群优化器相结合进行就业率预测

毕业生就业率的准确预测可以极大地帮助教育部门做出明智的决策以及大学调整教学计划。不幸的是，由于历史就业率数据存在波动，毕业生就业率的预测仍然是一个难题。本文提出了一种利用分数阶灰色GM(1,1sin)模型进行就业率预测的新方法，旨在减轻数据波动对预测精度的影响，并增加新数据在预测过程中的贡献。为了进一步减少预测误差，采用傅里叶级数对残差级数进行建模。所提出的模型称为傅立叶分数阶GM(1,1sin)模型(FFGMsin)，用于预测燕山大学2010年至2019年研究生的就业率。结果表明，所提出的方法可以获得比GM更准确的预测结果(1,1) 模型及其变体。© 2024 日本电气工程师协会。由 Wiley 期刊有限责任公司出版。