| 2,438 | 29 | 103 |
| 下载次数 | 被引频次 | 阅读次数 |
神经网络模型对大样本时间序列的拟合效果优于传统时间序列模型,但对于年度、月度、日度等低频时间序列的预测则难以发挥其优势。鉴于此,本文应用传统时间序列模型和神经网络模型,建立Holtwinters-BP组合模型,利用Holtwinters模型分别拟合各解释变量序列,利用BP模型拟合解释变量和自变量的非线性关系,基于某社交新闻类APP的日广告收入数据进行互联网企业广告收入预测研究。通过与循环神经网络(RNN)模型、长短期记忆神经网络(LSTM)模型等预测结果的对比发现:Holtwinters-BP组合模型的预测精度和稳定性更高;证明多维变量对于广告收入的显著影响,多变量模型的预测准确性高于单变量模型;构建的Holtwinters-BP组合模型对于低频数据预测有较好的有效性和适用性。
Abstract：The fitting effect of neural network model on large sample time series is often better than that of traditional time series models,but it may not be advantageous for the prediction of low-frequency time series such as annual,monthly and daily data. In view of this,we propose Holtwinters-BP combined model to forecast the advertisement income and expand the multi-dimensional explanatory variables,using daily advertisement income data from a social news type app. By comparing the prediction results with those of RNN and LSTM,it is found that Holtwinters-BP combined model has higher prediction accuracy and stability; explanatory variables have a significant impact on advertisement income,the prediction accuracy of the model with explanatory variables is higher than that of single variable model; it also verifies the validity and applicability of HoltwintersBP combined model for low-frequency data prediction.
[1]艾瑞咨询．2018年中国金融行业互联网广告投放将超过110亿[EB/OL]. http://report. iresearch. cn/content/2015/04/249243.shtml.
[2]中国互联网络信息中心．第42次中国互联网络发展状况统计报告[R]. 2018-08-20.
[3]中国互联网络信息中心．第43中国互联网络发展状况统计报告[R]. 2019-02-28
[4]方博，何朗．关于ARMA-BP神经网络组合模型的财政收入预测[J]．数学杂志，2015,35(3):709-713.
[5]谷建伟，周梅，李志涛，贾祥军，梁颖．基于数据挖掘的长短期记忆网络模型油井产量预测方法[J/OL]．特种油气藏，2019:1-8.
[6]蒋锋，张婷，周琰玲．基于Lasso-GRNN神经网络模型的地方财政收入预测[J]．统计与决策，2018,34(19):91-94.
[7]宋文玉．全球电视广告收入预测[J]．广播电视信息，2001(6):15.
[8]孙元，吕宁．地方财政一般预算收入预测模型及实证分析[J]．数量经济技术经济研究，2007(1):38-45.
[9]吴佳冕，杨桂元．基于IOWHA算子的中国货币供应量组合预测模型研究[J]．长春工程学院学报(自然科学版)，2017,18(4):121-125.
[10]吴晓峰，陈垚彤，赵桂彩．金融时间序列组合预测模型与方法[J]．中国市场，2019(9):47-50.
[11]姚琦．山西省机场广告收入影响因素分析及对策研究[D]．山西大学，2017.
[12]张延利，张德生．基于两个单项模型的组合预测模型构建方法[J]．统计与决策，2017(7):75-78.
[13]Bates J,Granger C. The Combination of Forecasts. Journal of the Operational Research Society,1969,20(4):451-468.
[14]Bengio S,Bengio Y,Cloutier J. Use of Genetic Programming for the Search of a New Learning Rule for Neural Networks[A].Proceedings of the First IEEE Conference on Evolutionary Computation. IEEE World Congress on Computational Intelligence,1994.
[15]Chen W. Phishing Detection Research Based on LSTM Recurrent Neural Network[A]. ICYCSEE Steering Committee. Abstracts of the4th International Conference of Pioneering Computer Scientists,Engineers and Educators(ICPCSEE 2018)Part I[C]. 2018:1.
[16]Dalymple D J. Using Box-Jenkins Techniques in Sales Forecasting[J]. Journal of Business Research,1978:133-145.
[17]Dietz S J,Kneringer P,Mayr G J,et al. Forecasting Low-Visibility Procedure States with Tree-Based Statistical Methods[J]. Pure and Applied Geophysics,2018(3):1-14.
[18]Hemanth K P,Patil B. Volatility Forecasting using Machine Learning and Time Series Techniques[J]. International Journal of Innovative Research in Computer and Communication Engineering,2015,9(3):84-92.
[19]Hochreiter S,Schmidhuber J. Long Short-term Memory[J]. Neural Computation,1997,9(8).
[20]Makridakis S,Winkler R. Averages of Forecasts:Some Empirical Results[J]. Management Science,1983,29(a):987-996.
[21]Stock J H,Watson M W. Combination Forecasts of Output Growth in a Seven-Country Data Set[J]. Journal of Forecasting,2004,23(6):405-430.
[22]Swapna G,Vinayakumar R,Soman K P. Diabetes Detection Using Deep Learning Algorithms[J]. ICT Express,2018(10).
[23]Thomas M,Fullerton JR,Roberto T,Lawrence W. Regional Econometric Income Forecast Accuracy[J]. Journal of Forecasting,2005,24:325-333.
[24]Ugurlu U,Oksuz L,Tas O. Electricity Price Forecasting Using Recurrent Neural Networks[J]. Energies,2018,5(11):1-23.
[25]Zhang P G. Time Series Forecasting Using a Hybrid ARIMA and Neural Network Model[J]. Neurocomputing,2003,50(1):159-175.
(1)收入数据通常为年度、月度、日度数据，相对于以分、秒为单位记录的高频时间序列而言，其样本量较少，这里统一称为低频数据。
基本信息:
DOI：10.19343/j.cnki.11-1302/c.2020.05.008
中图分类号:F713.8;F49
引用信息:
[1]吴翌琳,南金伶.互联网企业广告收入预测研究——基于低频数据的神经网络和时间序列组合模型[J].统计研究,2020,37(05):94-103.DOI:10.19343/j.cnki.11-1302/c.2020.05.008.
基金信息:
中国人民大学科学研究基金(中央高校基本科研业务费专项资金资助)项目(20XNL015)
2020-05-25
2020-05-25