基于改进贝叶斯模型的地铁备件需求量预测实证研究

Empirical Study on Demand Forecasting of Metro Spare Parts Based on Improved Bayesian Model

经典贝叶斯预测模型虽有效利用先验信息进行预测,但是有需要较多数据、操作复杂、结果误差值较大的特点。基于样本数据,通过实验的方式调整贝叶斯预测模型参数的计算方法,降低了数据要求,简化了贝叶斯模型操作的复杂程度,提高了预测的精确度,并将其应用于地铁备件需求预测的实例中,证明了模型的有效性,在一定程度上提高了地铁备件管理的效率。

In this paper, we adjusted the calculation method of the parameters in Bayesian forecasting model to lower its demand for data, simplify the operational requirement of the Bayesian model and improve its forecasting accuracy, and then applied the model to the empirical forecasting of metro spare parts to prove its validity.