基于奇异谱分析的我国航空客运量集成预测模型

针对时间序列包含噪声以及单一模型可能存在预测表现不稳定的问题,本文提出了一个基于奇异谱分析(SSA)的集成预测模型,并将其运用于我国年度航空客运量的预测中.首先,采用SSA方法对原始时间序列进行分解和重构,得到一个剔除噪声的时间序列,然后将其作为单整自回归移动平均模型(ARIMA)、支持向量回归模型(SVR)、Holt-Winters方法(HW)等单一模型的输入并进行预测,接着再采用加权平均集成预测方法(WA)将三种单一模型的预测结果进行综合集成.通过与各单一模型、基于经验模态分解方法(EMD)的模型以及简单平均集成预测方法(SA)的预测结果进行对比发现,本文所建模型具有较高的预测精度和较稳定的预测表现.最后,采用本文的模型对我国2014-2016年年度航空客运量进行了预测.     

A multiscale modeling approach incorporating ARIMA and anns for financial market volatility forecasting

The financial market volatility forecasting is regarded as a challenging task because of irreg ularity, high fluctuation, and noise. In this study, a multiscale ensemble forecasting model is proposed. The original financial series are decomposed firstly different scale components （i.e., approximation and details） using the maximum overlap discrete wavelet transform （MODWT）. The approximation is pre- dicted by a hybrid forecasting model that combines autoregressive integrated moving average （ARIMA） with feedforward neural network （FNN）. ARIMA model is used to generate a linear forecast, and then FNN is developed as a tool for nonlinear pattern recognition to correct the estimation error in ARIMA forecast. Moreover, details are predicted by Elman neural networks. Three weekly exchange rates data are collected to establish and validate the forecasting model. Empirical results demonstrate consistent better performance of the proposed approach.     

基于VMD的中国出口集装箱运价指数分析与组合预测

基于分解-重构-分项预测-集成思想,通过优选分解方法、优化重构方法、优选预测方法及合理选择集成方法等途径,构建了基于变分模态分解(VMD)的组合预测模型,对中国出口集装箱运价指数(CCFI)进行了预测,分析了CCFI波动特性及经济内涵.首先,选用VMD将运价指数序列分解为多个模态分量;其次,采用C值优化的FCM算法将模态分量重构为高、中、低频和趋势项,通过波动特性分析挖掘了重构项蕴含的短期市场不均衡因素、季节因素、重大事件及市场供需等经济内涵;再次,构建了基于数据特征分析的预测模型优选方法,进行了重构项预测;最后,将重构项预测值相加集成,分析了预测效果.实证结果表明,构建的组合模型预测效果优于BPNN、SVM、ARIMA等单一模型、EMD组合模型及未优化的VMD组合模型,较好地体现了CCFI外在波动特征与内在经济意义.     

CRUDE OIL PRICE FORECASTING WITH TEI@I METHODOLOGY

The difficulty in crude oil price forecasting, due to inherent complexity, has attracted much attention of academic researchers and business practitioners. Various methods have been tried to solve the problem of forecasting crude oil prices. However, all of the existing models of prediction can not meet practical needs. Very recently, Wang and Yu proposed a new methodology for handling complex systems-TEI@I methodology by means of a systematic integration of text mining, econometrics and intelligent techniques.Within the framework of TEI@I methodology, econometrical models are used to model the linear components of crude oil price time series (i.e., main trends) while nonlinear components of crude oil price time series (i.e., error terms) are modelled by using artificial neural network (ANN) models. In addition, the impact of irregular and infrequent future events on crude oil price is explored using web-based text mining (WTM) and rule-based expert systems (RES) techniques. Thus, a fully novel nonlinear integrated forecasting approach with error correction and judgmental adjustment is formulated to improve prediction performance within the framework of the TEI@I methodology. The proposed methodology and the novel forecasting approach are illustrated via an example.     

基于奇异谱分析的航空客运需求分析与分解集成预测模型

考虑到航空旅客运输需求影响因素复杂以及航空客运需求序列非线性非平稳等特征,本文提出了一个基于奇异谱分析(SSA)的航空客运需求分析与分解集成预测模型.需求分析阶段,首先使用SSA对航空客运需求序列进行有效分解,接着借助奇异熵理论,将序列重构为长期趋势项、中期市场波动项和短期噪声项;预测阶段,使用排列熵(PE)判断各重构序列复杂度的高低,并依据序列复杂度分别选择粒子群算法(PSO)和布谷鸟算法(CS)双优化的支持向量回归模型(SVR)或单整自回归移动平均模型(ARIMA)进行预测,结果表明,该分解集成预测模型较ARIMA、SVR等基准模型有着更好的预测性能.     

China’s energy consumption forecasting by GMDH based auto-regressive model

It is very significant for us to predict future energy consumption accurately. As for China’s energy consumption annual time series, the sample size is relatively small. This paper combines the traditional auto-regressive model with group method of data handling (GMDH) suitable for small sample prediction, and proposes a novel GMDH based auto-regressive (GAR) model. This model can finish the modeling process in self-organized manner, including finding the optimal complexity model, determining the optimal auto-regressive order and estimating model parameters. Further, four different external criteria are proposed and the corresponding four GAR models are constructed. The authors conduct empirical analysis on three energy consumption time series, including the total energy consumption, the total petroleum consumption and the total gas consumption. The results show that AS-GAR model has the best forecasting performance among the four GAR models, and it outperforms ARIMA model, BP neural network model, support vector regression model and GM (1, 1) model. Finally, the authors give the out of sample prediction of China’s energy consumption from 2014 to 2020 by AS-GAR model.     

Forecasting Container Throughput of Qingdao Port with a Hybrid Model

This paper proposes a hybrid forecasting method to forecast container throughput of Qingdao Port.To eliminate the influence of outliers,local outlier factor(lof) is extended to detect outliers in time series,and then different dummy variables are constructed to capture the effect of outliers based on domain knowledge.Next,a hybrid forecasting model combining projection pursuit regression(PPR) and genetic programming(GP) algorithm is proposed.Finally,the hybrid model is applied to forecasting container throughput of Qingdao Port and the results show that the proposed method significantly outperforms ANN,SARIMA,and PPR models.     

Forecasting China’s Foreign Trade Volume with a Kernel-Based Hybrid Econometric-Ai Ensemble Learning Approach

Due to the complexity of economic system and the interactive effects between all kinds of economic variables and foreign trade, it is not easy to predict foreign trade volume. However, the difficulty in predicting foreign trade volume is usually attributed to the limitation of many conventional forecasting models. To improve the prediction performance, the study proposes a novel kernel-based ensemble learning approach hybridizing econometric models and artificial intelligence （AI） models to predict China＇s foreign trade volume. In the proposed approach, an important econometric model, the co-integration-based error correction vector auto-regression （EC-VAR） model is first used to capture the impacts of all kinds of economic variables on Chinese foreign trade from a multivariate linear analysis perspective. Then an artificial neural network （ANN） based EC-VAR model is used to capture the nonlinear effects of economic variables on foreign trade from the nonlinear viewpoint. Subsequently, for incorporating the effects of irregular events on foreign trade, the text mining and expert＇s judgmental adjustments are also integrated into the nonlinear ANN-based EC-VAR model. Finally, all kinds of economic variables, the outputs of linear and nonlinear EC-VAR models and judgmental adjustment model are used as input variables of a typical kernel-based support vector regression （SVR） for ensemble prediction purpose. For illustration, the proposed kernel-based ensemble learning methodology hybridizing econometric techniques and AI methods is applied to China＇s foreign trade volume prediction problem. Experimental results reveal that the hybrid econometric-AI ensemble learning approach can significantly improve the prediction performance over other linear and nonlinear models listed in this study.     

基于小波分解的月径流预测校正模型研究

根据小波分析理论,建立了月径流序列的小波分解预测校正模型。该模型通过小波分解方法将月径流非平稳时间序列分解为多个细节信号序列和一个逼近信号序列,然后运用平稳时间序列的ARMA模型对各信号序列分别进行预测,最后再对各序列预测结果的和进行校正。以长江的宜昌站和寸滩站的月径流资料为例,分别采用ARMA模型、季节性ARIMA模型、BP神经网络模型以及所建立的小波分解预测校正模型进行模拟预测,并讨论了小波分解尺度数对小波分解预测校正模型的影响。结果表明,所建立的小波分解预测校正模型较其它模型具有更高的预测精度,并且尺度数对月径流序列模拟预测的效果没有显著的影响。     

数据特征驱动的火电产能过剩分解集成预测模型

有机融合数据特征驱动与多模态信息集成建模思想,构建了中国火电行业产能过剩组合预测方法和模型.首先识别火电产能过剩规模时序数据的本质和模式特征,发现其不仅具有非平稳,非线性特征,还呈现高复杂性和突变性;其次采用与数据特征相配的变分模态分解方法将时序数据分解,得到多个分量;然后识别各分量的数据特征,据此选择三次指数平滑-最小二乘支持向量机模型进行预测;最后集成各分量预测结果,得到火电产能过剩规模的最终预测结果.实证检验表明,所构建模型的预测水平精度,方向精度和稳定性均优于目前广泛使用的单一模型和其他组合预测模型.预测结果显示,2020-2022年中国火电产能过剩规模仍处于较高水平,呈先降后升趋势,且体制扭曲仍将是火电产能过剩的决定性因素.     

模糊软集合理论在税收组合预测中的应用

结合模糊软集合理论建立税收收入的组合预测模型,根据税收收入的特点,代表性地选择了Elman回归神经网络模型、含政策虚拟变量的自回归模型、ARIMA(1,1,1)的时间序列模型、多因素SVM回归模型这四种模型作为组合预测中的单一模型,并以1980年到2008年的税收收入等相关数据为背景进行了说明和分析.结果表明该组合预测模型能有效减小预测误差,为税收工作实践提供了一个应用研究工具,并推广和丰富了软集合理论在税收经济模型研制中的实际应用.     

Artificial Neural Network for Combining Forecasts

ＡｒｔｉｆｉｃｉａｌＮｅｕｒａｌＮｅｔｗｏｒｋｆｏｒＣｏｍｂｉｎｉｎｇＦｏｒｅｃａｓｔｓ￥ＳｈａｎｍｉｎｇＳｈｉ，ＬｉＤ．Ｘｕ＆ＢａｏＬｉｕ（ＤｅｐａｒｔｍｅｎｔｏｆＣｏｍｐｕｔｅｒＳｃｉｅｎｃｅ，ＵｎｉｖｅｒｓｉｔｙｏｆＣｏｌｏｒａｄｏａｔＢｏｕｌ...     

神经网络模型用于多变量综合预测

本文研究神经网络用于多变量时间序列预测的原理与方法，提出组合多种信息的综合预测方法。以股票交易为例，用神经网络组合各类信息，运用信心股价理论对中国股市的发展进行跟踪预测。在此基础上进一步从信息利用的角度说明了神经网络预测方法的特点。结果表明，神经网络模型用于多变王时间序列预测，其精度和趋势均较统计方法有所提高；神经网络综合预测，对中短期股票价格的预测，有实用价值。     

基于串联灰色神经网络的电力负荷预测方法

为了提高电力负荷预测的精度,分析现有人工神经网络和灰色预测方法各自的优缺点,将二者相结合提出了一种串联灰色神经网络预测方法.新方法利用灰色预测中的累加生成运算对原始数据进行变换,从而得到规律性较强的累加数据,便于神经网络进行建模和训练,同时避免了灰色预测方法存在的理论误差.最后实际算例证明了方法的有效性.方法适用于中长期负荷预测.     

基于乘积ARIMA模型的产品不确定性需求预测

为对不确定的市场需求进行有效预测,可运用SAS系统中的TimeSeriesForecastingSystem,基于乘积求和自回归平均滑动模型(ARIMA),对产品销售的时间序列数据进行预测。先对原始数据或变换后的数据作简单差分或季节差分,把时间序列化为平稳的时间序列,进行参数的初估计,然后进行多次拟合并最终确定模型,根据SAS估计结果,可得到预测偏差。对预测结果在营销管理和供应链管理中的应用进行了分析,取得了较好的效果。     

Time Series Neural Network Forecasting Methods

ＴｉｍｅＳｅｒｉｅｓＮｅｕｒａｌＮｅｔｗｏｒｋＦｏｒｅｃａｓｔｉｎｇＭｅｔｈｏｄｓＷＥＮＸｉｎｈｕｉ；ＣＨＥＮＫｅｉｚｈｏｕ（ＴｈｅＣｅｎｔｌａｌｏｆＮｅｕｒａｌＮｅｔｗｏｌｋ，Ｘｉ＇ｄｉａｎＵｎｉｖｅｒｓｉｔｙ，Ｘｉａｎ７１００７１，Ｃｈｉｎａ）Ａ...     

基于ARIMA模型的短时交通流实时自适应预测

实时、准确的短时交通流量预测是智能交通系统(ITS)中的一个关键问题。基于采用ARIMA(P，d，0)模型结构的时间序列分析方法，提出一种短时交通流实时自适应预测算法。在该算法中采用带遗忘因子的递推最小二乘方法进行参数估计，采用基于线性最小方差预报原理的Astrom预报算法进行预报。针对大量实测数据进行仿真实验，结果表明：减小遗忘因子可以提高一步预测的性能。此外，将该算法分别应用于工作日和双休日的数据时，仿真实验都取得了较好的预测效果，说明该算法对不同交通流状况具有较好的适应性。     

基于DLENN模型的沈阳地区PM2.5浓度预测

受到重工业发展规模、北温带季风气候、秋冬季燃煤取暖、机动车拥堵状况以及微观气象条件等各种因素影响,沈阳地区PM2.5浓度变化具有趋势性、周期性及随机性特征.针对上述三种特征,论文构建了一种集成双向长短期记忆网络的神经网络预测模型DLENN(Double-LSTM Ensemble Neural Network),内含的...     

金融危机背景下的人民币汇率预测

在为金融危机期间人民币汇率的波动提供一种有效的预测方法.在利用替代数据方法检验和判别汇率系统具有非线性结构的基础上,识别了各具体汇率序列的最优滞后期组合,并分别采用了多层感知机(MLP)和层反馈网络(RNN2)结构构建同质神经网络模型,从三个方面对比分析了模型群在不同参数条件下的预测效果. 研究发现,根据不同序列的具体特征,各神经网络模型在不同自由度下的4个预测期限内的预测性能存在较明显的差异.同时,包含层反馈过程的RNN2模型在描述与预测人民币汇率的波动方面表现出很强的能力.此外, 还分析并解释了产生上述结果的原因,并为4种人民币汇率波动序列甄选出了相应的最优预测模型.     

A KELM-Based Ensemble Learning Approach for Exchange Rate Forecasting

In this paper, a KELM-based ensemble learning approach, integrating Granger causality test, grey relational analysis and KELM(Kernel Extreme Learning Machine), is proposed for the exchange rate forecasting. The study uses a set of sixteen macroeconomic variables including, import,export, foreign exchange reserves, etc. Furthermore, the selected variables are ranked and then three of them, which have the highest degrees of relevance with the exchange rate, are filtered out by Granger causality test and the grey relational analysis, to represent the domestic situation. Then, based on the domestic situation, KELM is utilized for medium-term RMB/USD forecasting. The empirical results show that the proposed KELM-based ensemble learning approach outperforms all other benchmark models in different forecasting horizons, which implies that the KELM-based ensemble learning approach is a powerful learning approach for exchange rates forecasting.