【摘要】 本研究利用ARIMA模型能够通过对有限时间长度的历史销售数据分析而实现精度较高的短期性预测的特点,针对珍贵木材首饰零售企业的具有代表性的、随机性需求的首饰进行数据的序列平稳化过程处理,而后进行模型的识别、参数估计和假设检验,最后进行预测,以检验ARIMA模型预测的准确度。结果表明,针对该类木材首饰而建立的ARIMA预测模型,其模型拟合优度和白噪声等参数值均具有统计学意义,预测结果也证明其短期预测的准确度较高。而未来控制库存的核心是对商品将来的需求进行有效的预测,因此利用ARIMA模型短期预测精度高的优点,可以为木材首饰零售行业短期库存控制提供决策和依据。更多还原

【Abstract】 The ARIMA model can analyze historical sales data with limited length to achieve higher accuracy of short- term prediction. In this paper,based on the wood jewelry sales data with representativeness and random demand in a typical precious wood jewelry retailer,a stabilized treatment was conducted on the time series,followed by model identification,parameter estimation,hypothesis testing,and prediction in order to test the prediction accuracy of ARIMA model. The results showed that for the established ARIMA model,the model goodness of fit and white noise were statistically significant,and the results also demonstrated the higher forecast accuracy in the short- term. The future core of inventory control is to effectively predict the future demand of commodities.Thus,taking the advantage of high short- term forecasting accuracy of ARIMA model can provide decision- making and basis for wood jewelry retail industry in short- term inventory control.更多还原