On a constrained mixture vector autoregressive model

A mixture vector autoregressive model has recently been introduced to the literature. Although this model is a promising candidate for nonlinear multiple time series modeling, high dimensionality of the parameters and lack of method for computing the standard errors of estimates limit its application to real data. The contribution of this paper is threefold. First, a form of parameter constraints is introduced with an efficient EM algorithm for estimation. Second, an accurate method for computing standard errors is presented for the model with and without parameter constraints. Lastly, a hypothesis-testing approach based on likelihood ratio tests is proposed, which aids in the selection of unnecessary parameters and leads to the greater efficiency at the estimation. A case study employing U.S. Treasury constant maturity rates illustrates the applicability of the mixture vector autoregressive model with parameter constraints, and the importance of using a reliable method to compute standard errors.     

基于时间序列分类任务的智能电能表负荷监测技术研究

随着深度学习模型在非侵入式负荷监测领域的应用,负荷识别与分解的能力得到了显著提升。但是多数方法仍然存在模型训练效率低下、分解精度不足以及模型不易推广的问题。针对上述问题,文章利用基于时间序列分类任务的卷积神经网络对非侵入式负荷监测框架进行了研究,并提出了相应的负荷识别与分解方法。通过对比实验证明,在UK-DALE数据集上,基于时间序列分类任务的卷积神经网络提升了洗碗机4.3%的识别准确率和19%的识别精度,降低了21.3%负荷分解过程的均方误差;在REDD数据集上,模型对于洗衣机的识别准确率、精度和F1值均有所提升,特别是召回率提高了24.3%,同时在负荷分解的过程中,模型降低了15.8%的均方误差。因此,与其它神经网络模型相比,基于时间序列分类任务的卷积神经网络具有更稳定的负荷识别与分解性能。     

基于抗差估计的灰色模型与时间序列组合模型及其在变形预测中的应用

文章讨论的方法是在时序分析建模的基础上,运用了GM(1,1)来提取趋势项,然后对残差进行时序建模,形成灰色与时序的组合模型,在建模中的参数估计方法,采用抗差估计来取代常规的最小二乘估计法,来达到抵抗粗差的目的,从实例计算来看,效果比较好。     

基于时序分析的拉深件成型裂纹的特征参数提取

研究金属拉深制件声发射信号特征参数的提取对判断成型制件的质量尤为重要。研究采用时序分析和MATLAB对经过小波包分解的声发射信号进行特征参数的提取。在此过程中利用时间序列的自相关系数和偏相关系数的拖尾性及截尾性来判断模型的类型;采用FPE (final prediction error,最终预测误差0029准则对模型进行定阶;以最小二乘估计法对模型的参数进行估计。最后,根据所建立的自回归谱模型提取声发射信号的特征参数。研究结果是通过此方法成功提取到了金属拉深制件声发射信号的特征参数及其分布图,为成型制件质量的判断提供了有利的依据。     

基于PSO-SVM的大坝渗流监测时间序列非线性预报模型

大坝渗流监测分析是大坝安全监控的重要内容，预测分析的难点之一在于渗流监测数据往往具有复杂的非线性特点。本文充分利用支持向量机的结构风险最小化与粒子群算法快速全局优化的特点，采用粒子群算法快速优化支持向量机的模型参数，通过该模型对非线性监测数据进行拟合，建立了基于PSO_SVM的大坝渗流监测的时间序列非线性预报模型。本模型应用于隔河岩水电站的坝基渗流量的预测，计算结果与实际监测值吻合良好。     

基于多变量监测时序的冲击地压复杂性分析

论文基于多变量时间序列相空间重构来计算数据的关联维数,以研究冲击地压监测数据的复杂程度。考虑到冲击地压监测数据含有噪声而且长度有限,对传统G-P算法进行了扩展改进,给出了改进算法求解多变量时间序列的关联维数的原理,并用于Lorenz混沌系统检验了改进算法的有效性。收集了不同冲击情况下多种监测类型的冲击地压时间序列数据,用改进G-P算法求解这些监测数据的关联维数值。研究结果表明:冲击地压监测数据具有混沌特性,而且数据关联维数越大,复杂程度越高,对应矿井的冲击破坏性越强。这为基于混沌理论预测冲击危险性提供了新方法和依据。     

PREDICTING INFLUENZA WAVES WITH HEALTH INSURANCE DATA

In recent years, many of the most developed countries have started to create influenza surveillance systems, because influenza still is very costly, not just concerning the health systems, but also economically. In most of these systems statistical methods are applied, unfortunately with rather moderate results. In contrast to statistical methods case-based reasoning explicitly uses former episodes. Because we already successfully applied our prognostic method, which combines case-based reasoning (CBR) with temporal abstraction, to kidney functions, we use it again to forecast influenza. Because health centers collect extensive laboratory data but their availability is usually delayed for at least two weeks, we use quickly available data from the main German health insurance scheme. In this article, we propose the use of CBR for influenza forecast and we show promising results.     

Efficient order selection algorithms for integer-valued ARMA processes

Abstract.  We consider the problem of model (order) selection for integer-valued autoregressive moving-average (INARMA) processes. A very efficient reversible jump Markov chain Monte Carlo (RJMCMC) algorithm is constructed for moving between INARMA processes of different orders. An alternative in the form of the EM algorithm is given for determining the order of an integer-valued autoregressive (INAR) process. Both algorithms are successfully applied to both simulated and real data sets.     

Use of modular architectures for time series prediction

Recently, a lot of papers have been published in the field of time series prediction using connectionist models. Nevertheless we think that one of the major problem with is rarely treated in the literature is related to the choice of input parameters (embedding dimension and delay). In this paper, we propose two modular approaches to this problem and apply them to a sunspot-related time series. Experimental results are then compared to a single multi-layer perceptron in order to estimate performances of these models.     

时间序列中的分形分析和参数提取

本文我们主要讨论拓扑维为一维的时间序列的分形分析和参数提取，着重论述分维数的估计，介绍了几种常用方法，并提出了一种基于有限长度的方差估计法，用实验分析了这些计算方法的准确性、平稳性、计算复杂度、受噪声和信号长度的影响等方面的性能，得到一些有意义的结论，实验结果表明，我们提出的方法在某些方面具有优点。