基于时间序列模型和情感分析的情感趋势预测

为解决传统情感分析方法无法对公众未来情感走势变化有效预测的问题,提出一种将时间序列模型与情感分析相结合的情感趋势预测方法.采用深度学习模型对股市论坛实时评论信息进行情感分类,统计固定时间单位的情感值,构建情感值时间序列,提出ARIMA-GARCH时间序列模型,对情感值时间序列进行建模分析,预测投资者的情感走势.实验结果表明,该方法对于情感趋势的预测结果合理,误差较小.同时,发现投资者情感趋势与股市涨跌幅走势相似,为投资决策提供了参考.     

基于ARIMA和LSTM混合模型的时间序列预测

由于现实中的时间序列通常同时具有线性和非线性特征,传统ARIMA模型在时间序列建模中常表现出一定局限性.对此,提出基于ARIMA和LSTM混合模型进行时间序列预测.应用线性ARIMA模型进行时间序列预测,用支持向量回归(SVR)模型对误差序列进行预测,采用深度LSTM模型对ARIMA模型和SVR模型的预测结果组合,并将...     

基于ARIMA-LSSVM混合模型的犯罪时间序列预测

对犯罪时间序列的预测对帮助公安部门更好地掌握犯罪动态,实现智能犯罪发现具有重大意义。针对犯罪时间序列预测的计算需求,结合真实犯罪数据集,提出了ARIMA-LSSVM混合模型。该模型通过ARIMA预测出时间序列的线性部分,通过PSO优化的LSSVM模型预测非线性部分,以对序列进行充分拟合,最后通过混合算法计算最终结果。使用此混合模型达到了精准的预测效果,证明了模型的有效性。     

基于时间序列线性拟合的色谱数据压缩方法

从石油录井色谱数据应用的实际需求出发,提出一种新的时间序列分段拟合算法。该算法通过一次扫描数据,根据中线距离阈值和非单调序列中极值点保持时间段阈值两个约束条件,选择反映序列趋势变化的关键点,然后线性拟合时间序列。实验结果表明该算法能够在保持原始序列主要形态的同时剔除噪音干扰,精确定位单调序列中的突变转折点,发现序列中的尖峰状态。     

用于多元时间序列预测的自适应频域模型

近年来,学术和工业领域对时间序列数据的研究热潮不断增长,但其中蕴含的频率信息仍缺乏有效的建模.研究发现,时间序列预测依赖于不同的频率模式,为未来的趋势预测提供有用的线索:短期的序列预测更多依赖于高频分量,而长期预测则更多关注低频数据.为更好地挖掘时间序列的多频模式,提出了一个多特征自适应频域预测模型MAFD.该模型分为两个阶段:在第一阶段中,模型通过XGBoost算法对输入向量进行重要性度量,选择高重要性特征;在第二阶段,模型将时间序列的频率特征提取和目标序列的频域建模集成到一起,并根据时间序列对频率模式的依赖特点构建一个端到端的预测网络.MAFD的创新性体现在预测网络能够根据输入序列的动态演变自动关注不同的频率分量,从而揭示时间序列的多频模式,强化模型的学习能力.采用4种不同领域的数据集对模型进行了性能验证,实验结果表明,与现有经典的预测模型相比,MAFD具有更高的准确性和更小的滞后性.     

基于趋势转折点的时间序列分段线性表示

分段线性表示是时间序列降维的有效方法。在总结分析序列趋势变化特点的基础上,提出了一种基于趋势转折点的时间序列分段线性表示算法。首先定义了趋势转折点作为时间序列分段点的备选集,以点到区域的距离度量趋势转折点的重要性,再根据给定的阈值选择重要趋势转折点作为分段点,对时间序列进行分段线性表示。通过与其他6种方法进行实验比较,结果表明：所提方法在具有较好的拟合质量和适应能力以及对转折点明显的序列,都表现出较强的抗噪声干扰能力。     

短记忆性时间序列的建模分析及预测

AR模型作为时间序列模型的一种,由于其参数估计和定阶简单而广泛用于系统辨识.在多维AR序列的最小二乘建模的基础上,结合Kalman滤波算法,推导了应用Kalman滤波技术的多维AR序列参数估计方法.该算法无需保存历史数据,可对AR模型的估计参数进行实时的修正.在确定AR模型阶数时,提出了两步F检验法.选取上证某A股收益序列作为样本,利用时间序列相关性分析对该算法的有效性进行验证;对时间序列的RMSE和MAD指标进行比较,结果表明该算法大大减少了建模过程中的计算工作量,并具有较好的预测性.     

基于多元线性自回归模型的流量预测

为使流量预测模型具有自适应性和相关性,以时间点为基础进行建模,结合时间序列与流量序列,引入多元线性自回归(AR)思想进行参数估算,对多次估算所得参数值建立指数加权移动平均数模型进行二次估算,在此基础上,建立多元线性自回归模型。实验结果证明,与AR模型、ARMA模型相比,基于多元线性AR模型的预测结果更准确。     

基于未知输入观测器的非线性时间序列故障预报

将未知非线性系统的输出作为时间序列并进行空间重构,针对得到的离散线性时变系统,提出了基于未知输入观测器的预测新方法.以实时拟合时间序列的线性AR模型作为时变系统的已知线性部分,将拟合误差作为时变系统的未知输入,实现了对非线性时间序列的一步预测.再利用递推预测的方法,将一步预测推广到N步预测,同时证明了该方法的预测误差有界.通过未知输入的预测值和状态的预测误差的变化可以方便地判断故障的发生,实现故障预报.仿真结果证明了方法的有效性.     

改进嵌入维数和时间延迟计算的GP预测算法

改进了混沌系统中的两个重要特征量:嵌入维数和时间延迟的计算,根据计算得出的上述两个参数重构相空间;然后在相空间中作轨迹的线性拟合,选择轨迹中的最近邻点作一次性的预测.提出的算法在相空间中很好地把轨迹的线性拟合与最近邻方法结合起来,解决了现有的时间序列分析和预测算法中主观性太强的缺点,通过对话务量时间序列和太阳黑子时间序列的验证,与其它算法相比,该算法的分析结果稳定而准确、预测精度高、运行时间比较短.     

基于定量递归联合熵特征重构的缓冲区流量预测算法

对网络基站缓冲区的短时网络流量的准确预测是缓解和控制拥堵的关键.基站缓冲区的短时网络流量时间序列具有非线性混沌特征,其自相关特性较弱,而传统方法采用线性时间序列分析方法没能有效挖掘流量序列的非线性特征信息,流量序列预测性能不好.提出了一种基于非线性时间序列分析的定量递归联合熵特征重构网络基站缓冲区的短时网络流量预测算法,该算法提取流量序列的定量递归联合熵特征,并对特征序列进行相空间重构;把网络流量信号模型进行高维映射,在高维相空间对短时网络流量序列进行定量递归分析;采用自相关特征奇异分解对流量序列进行聚合后的线性叠加,采用平均互信息算法和虚假最近邻点算法计算最佳时延参数和最小嵌入维;进行插值拟合形成时频分析特征分流控制,实现对网络流量的预测.仿真结果表明,该算法预测精度较高,稳定性较好,预测偏差较传统方法低,具有较好的应用价值.     

基于改进的RBF神经网络的人民币汇率预测研究

针对RBF神经网络分段算法中对近似线性时间序列数据预测误差较大这一不足,在原有RBF神经网络模型基础上提出了一种改进算法。该算法以分段取中心值为基础,优化原算法中径向基函数中心点值的确定,提高了对近似线性时间序列数据预测的准确度。通过对近两年美元兑人民币汇率数据的预测测试,表明改进算法在预测准确性比原算法有较大提高。     

Predicting Chaotic Time Series Using Neural and Neurofuzzy Models: A Comparative Study

The prediction accuracy and generalization ability of neural/neurofuzzy models for chaotic time series prediction highly depends on employed network model as well as learning algorithm. In this study, several neural and neurofuzzy models with different learning algorithms are examined for prediction of several benchmark chaotic systems and time series. The prediction performance of locally linear neurofuzzy models with recently developed Locally Linear Model Tree (LoLiMoT) learning algorithm is compared with that of Radial Basis Function (RBF) neural network with Orthogonal Least Squares (OLS) learning algorithm, MultiLayer Perceptron neural network with error back-propagation learning algorithm, and Adaptive Network based Fuzzy Inference System. Particularly, cross validation techniques based on the evaluation of error indices on multiple validation sets is utilized to optimize the number of neurons and to prevent over fitting in the incremental learning algorithms. To make a fair comparison between neural and neurofuzzy models, they are compared at their best structure based on their prediction accuracy, generalization, and computational complexity. The experiments are basically designed to analyze the generalization capability and accuracy of the learning techniques when dealing with limited number of training samples from deterministic chaotic time series, but the effect of noise on the performance of the techniques is also considered. Various chaotic systems and time series including Lorenz system, Mackey-Glass chaotic equation, Henon map, AE geomagnetic activity index, and sunspot numbers are examined as case studies. The obtained results indicate the superior performance of incremental learning algorithms and their respective networks, such as, OLS for RBF network and LoLiMoT for locally linear neurofuzzy model.     

一种非线性自适应混沌时间序列预测方法

基函数线性叠加的混沌时间序列预测算法不具有动态特性和明确的物理意义。改进的策略使用与混沌序列的非高斯特性相联系的函数作为基函数,使其能解释为表征混沌序列的高阶统计特性。同时,在算法中引入非线性反馈环节,使其具有了动态特性。数值仿真表明,以之为基础的自适应预测算法在一步预测性能和长期预测能力方面都优于常用的线性预测方法和已有的自适应预测算法。     

Efficient financial time series prediction with evolutionary virtual data position exploration

Prediction of stock index remains a challenging task of the financial time series prediction process. Random fluctuations in the stock index make it difficult to predict. Usually the time series prediction is based on the observations of past trend over a period of time. In general, the curve the time series data follows has a linear part and a non-linear part. Prediction of the linear part with past history is not a difficult task, but the prediction of non linear segments is difficult. Though different non-linear prediction models are in use, but their prediction accuracy does not improve beyond a certain level. It is observed that close enough data positions are more informative where as far away data positions mislead prediction of such non linear segments. Apart from the existing data positions, exploration of few more close enough data positions enhance the prediction accuracy of the non-linear segments significantly. In this study, an evolutionary virtual data position (EVDP) exploration method for financial time series is proposed. It uses multilayer perceptron and genetic algorithm to build this model. Performance of the proposed model is compared with three deterministic methods such as linear, Lagrange and Taylor interpolation as well as two stochastic methods such as Uniform and Gaussian method. Ten different stock indices from across the globe are used for this experiment and it is observed that in majority of the cases performance of the proposed EVDP exploration method is better. Some stylized facts exhibited by the financial time series are also documented.

     

基于距离和DF-RLS的时间序列异常检测

为能同时检测时间序列中的附加异常和革新异常,改进自回归模型,提出距离因子递推最小二乘(DF-RLS)线性预测算法。在此基础上,给出一种基于距离和DF-RLS的联合异常检测方法——DDR-OD。实验结果表明,与当前其他时间序列异常检测方法相比,DDR-OD的检测效果较优。     

基于时间序列的支持向量机在股票预测中的应用

由于股票预测是不确定、非线性、非平稳的时间序列问题,传统的方法往往难以取得满意的预测效果。本文提出一种基于时间序列的支持向量机（SVM）股票预测方法。利用沙河股份的股票数据,建立股票收盘价回归预测模型,该模型克服了传统时间序列预测模型仅局限于线性系统的情况。实验结果表明,该方法比神经网络方法以及时间序列方法的预测精度更高,可以很好的应用某些非线性时间序列的预测中。     

基于参考点预测的动态多目标优化算法

为了快速跟踪动态多目标优化问题变化的Pareto前沿,本文提出一种基于参考点预测策略的动态多目标优化算法（PDMOP）.该算法对关联到相同参考点的个体建立时间序列,并对这些时间序列通过线性回归模型预测新环境下种群.同时,将历史时刻的预测误差反馈到当前预测中来提高预测的准确性,并在每个预测的个体上加入扰动来增加初始种群多样性,从而能够加快算法在新环境下的收敛速度.通过4个标准测试函数对该算法测试,并和两个现有算法对比分析,结果表明所提算法在处理动态多目标优化问题时能够保持良好的性能.     

Extracting the main patterns of natural time series for long-term neurofuzzy prediction

A combination of singular spectrum analysis and locally linear neurofuzzy modeling technique is proposed to make accurate long-term prediction of natural phenomena. The principal components (PCs) obtained from spectral analysis have narrow band frequency spectra and definite linear or nonlinear trends and periodic patterns; hence they are predictable in large prediction horizon. The incremental learning algorithm initiates a model for each of the components as an optimal linear least squares estimation, and adds the nonlinear neurons if they help to reduce error indices over training and validation sets. Therefore, the algorithm automatically constructs the best linear or nonlinear model for each of the PCs to achieve maximum generalization, and the long-term prediction of the original time series is obtained by recombining the predicted components. The proposed method has been primarily tested in long-term prediction of some well-known nonlinear time series obtained from Mackey–Glass, Lorenz, and Ikeda map chaotic systems, and the results have been compared to the predictions made by multi-layered perceptron (MLP) and radial basis functions (RBF) networks. As a real world case study, the method has been applied to the long-term prediction of solar activity where the results have been compared to the long-term predictions of physical precursor and solar dynamo methods.