Time Series Segmentation Using a Novel Adaptive Eigendecomposition Algorithm

In this paper we present a new technique for time series segmentation built around a fast principal component analysis (PCA) algorithm that is on-line and stable. The traditional Generalized Likelihood Ratio Test (GLRT) has been used to solve the segmentation problem, but this has enormous limitations in terms of complexity and speed. Newer methods use gated experts and mixture models to detect transitions in time series. These techniques perform better than GLRT, but most of them require extensive training of relatively large neural networks. The segmentation method discussed in this paper is based on a novel idea that involves solving the generalized eigendecomposition of two consecutive windowed time series and can be formulated as a two-step PCA. Thus, the performance of our segmentation technique mainly depends on the efficiency of the PCA algorithm. Most of the existing techniques for PCA are based on gradient search procedures that are slow and they also suffer from convergence problems. The PCA algorithm presented in this paper is both online, and is proven to converge faster than the current methods.     

Financial Times Series Forecasting of Clustered Stocks

Mobile Networks and Applications - Predicting the stock market is a widely studied field, either due to the curiosity in finding an explanation for the behavior of financial assets or for financial...     

An Automatic Forecasting Method for Time Series

An automatic forecasting method is pro-posed concerning automation problem in the field of lin-ear time series forecasting. The method is on the basis of econometric theory and overcomes the difficulty to mine and forecast automatically with econometric models. The proposed algorithm is divided into 4 stages, i.e. preprocess-ing, unit root testing and stationary processing, modeling, and ultimately forecasting. Future values and trends would be estimated and forecasted precisely through the 4 stages of the algorithm according to input data without man-ual intervention. Experimental comparisons were made be-tween the proposed algorithm and the 2 data driven fore-casting algorithms, i.e. moving average method and Holt exponential smoothing method. It was demonstrated with the experimental results that automatic forecasting is fea-sible utilizing the proposed algorithm and higher accuracy can be acquired than these 2 data driven-based methods.     

Adaptive Channel Normalization Based on Infomax Algorithm for Robust Speech Recognition

This paper proposes a new data‐driven method for high‐pass approaches, which suppresses slow‐varying noise components. Conventional high‐pass approaches are based on the idea of decorrelating the feature vector sequence, and are trying for adaptability to various conditions. The proposed method is based on temporal local decorrelation using the information‐maximization theory for each utterance. This is performed on an utterance‐by‐utterance basis, which provides an adaptive channel normalization filter for each condition. The performance of the proposed method is evaluated by isolated‐word recognition experiments with channel distortion. Experimental results show that the proposed method yields outstanding improvement for channel‐distorted speech recognition.     

短时序列预测的新方法

We propose a procedure to forecast short time series with stable seasonal pattern. This new method is motivated by the observations that short time series arise in many situations for the fierce competition. The quantity to be predicted is a yearly accumulation assuming that the partially accumulated data within the year are available. A simple model is proposed to describe the relationship between the yearly accumulation and partial accumulation and analytic results are obtained for both the point prediction and the predicative distribution. A comparison will be conducted between this model and traditional time series forecasting model with data from telecommunication industry. This method works better than the traditional models when only small amount of data are available. It can also be applied to forecast individual observations with a proper disaggregation algorithm.     

一种启发式变阶直觉模糊时间序列预测模型

论文针对已有高阶模糊时间序列模型在预测精度和预测范围上的限制,结合直觉模糊集理论,提出一种启发式变阶直觉模糊时间序列预测模型。模型首先应用直接模糊聚类算法对论域进行非等分划分;然后,针对直觉模糊时间序列的数据特性,改进现有直觉模糊集隶属度和非隶属度函数的建立方法;最后,采用阶数随序列实时变化的高阶预测规则进行预测,并将历史数据发展趋势的启发知识引入解模糊过程,使模型的预测范围得到扩展。在Alabama大学入学人数和北京市日均气温两组数据集上分别与典型方法进行对比实验,结果表明该模型有效克服了传统模型的缺点,拥有较高的预测精度,证明了模型的有效性和优越性。     

基于小波分解的某些非平稳时间序列预测方法

提出一种时间序列预测方法,称为小波预测方法.通过小波分解可以将某些非平稳时间序列分解成多层近似意义上的平稳时间序列,然后采用自回归模型对分解后的时间序列进行预测,从而得到原始时间序列的预测值.对年平均太阳黑子数的预测结果表明,该方法比传统的时间序列预测方法和神经网络预测方法的预测精度高,可以很好地应用于某些非平稳时间序列的预测中.     

Self-adaptive Processing and Forecasting Algorithm for Univariate Linear Time Series

As the Box-Jenkins method could not grasp the non-stationary characteristics of time series exactly, nor identify the optimal forecasting model order quickly and precisely, a self-adaptive processing and forecasting algorithm for univariate linear time series is proposed. A self-adaptive series characteristic test framework which employs varieties of statistic tests is constructed to solve the problem of inaccurate identification and inadequate processing for non-stationary characteristics of time series. To achieve favorable forecasts, an optimal forecasting model building algorithm combined with model filter and candidate model pool is proposed, in which a univariate linear time series forecasting model is built. Experimental data demonstrates that the proposed algorithm outperforms the comparativemethod in all forecasting performance statistics.     

基于自适应增量学习的时间序列模糊聚类算法

针对现存可用于时间序列的增量式模糊聚类算法往往需要设置多个控制参数的问题,本文提出了一种基于自适应增量学习的时间序列模糊聚类算法.该算法首先继承上一次聚类得到的簇结构信息以初始化当前聚类进程,然后在无需设置参数的情况下自适应地搜索当前数据块中的离群样本,并自动从离群样本创建新簇,最后检查空簇识别标识确定是否需要移除部分簇以保证后续聚类过程的效率.实验结果表明所提算法对等长和不等长时间序列均具有良好的聚类准确性及运行效率.     

分区复杂性在时间序列分析中的应用

本文对KC复杂性,C₁、C₂复杂性等常见的时间序列分析中所用的方法进行了比较,在此基础上提出一种新算法——分区复杂性,同时将其应用在EEG时间序列分析、处理中,以检验它的有效性.     

采用自适应观测器的共轴多旋翼无人机鲁棒控制

针对共轴多旋翼无人机输入受限和不确定性问题,提出了一种基于自适应观测器的鲁棒控制方法.首先建立了包含不确定性的共轴十二旋翼无人机位置和姿态数学模型,在位置回路鲁棒控制律的设计中,引入辅助观测器来补偿输入受限的影响;然后在姿态回路鲁棒控制律的设计中,引入自适应观测器来抑制不确定性的影响,从而实现共轴多旋翼无人机的全局渐近稳定.仿真结果表明:提出的鲁棒控制律具有更优的稳定性、准确性和快速性,位置和姿态的最大跟踪误差分别仅为0.05 m和0.2°,能够在更短的时间内实现对共轴多旋翼无人机的鲁棒控制.     

一种预测频段占用度的时间序列分析方法

提出了一种基于实际测试数据的频段占用度短期预测方法,该方法首次将经济学中常用的自回归移动平均模型(ARIMA)用于频段占用度的预测,并分析给出了判定模型参数和分析结果。短期预测结果与实际测试结果证明了这种预测方法可以为无线电频率主管部门提供更为精确地掌握频率资源的使用及发展变换情况,为制定政策和指配频率提供基础。     

Robust Deep Age Estimation Method Using Artificially Generated Image Set

Human age estimation is one of the key factors in the field of Human–Robot Interaction/Human–Computer Interaction (HRI/HCI). Owing to the development of deep‐learning technologies, age recognition has recently been attempted. In general, however, deep learning techniques require a large‐scale database, and for age learning with variations, a conventional database is insufficient. For this reason, we propose an age estimation method using artificially generated data. Image data are artificially generated through 3D information, thus solving the problem of shortage of training data, and helping with the training of the deep‐learning technique. Augmentation using 3D has advantages over 2D because it creates new images with more information. We use a deep architecture as a pre‐trained model, and improve the estimation capacity using artificially augmented training images. The deep architecture can outperform traditional estimation methods, and the improved method showed increased reliability. We have achieved state‐of‐the‐art performance using the proposed method in the Morph‐II dataset and have proven that the proposed method can be used effectively using the Adience dataset.     

一种在线时间序列预测的核自适应滤波器向量处理器

针对信息物理融合系统中的在线时间序列预测问题,该文选择计算复杂度低且具有自适应特点的核自适应滤波器(Kernel Adaptive Filter, KAF)方法与FPGA计算系统相结合,提出一种基于FPGA的KAF向量处理器解决思路。通过多路并行、多级流水线技术提高了处理器的计算速度,降低了功耗和计算延迟,并采用微码编程提高了设计的通用性和可扩展性。该文基于该向量处理器实现了经典的KAF方法,实验表明,在满足计算精度要求的前提下,该向量处理器与CPU相比,最高可获得22倍计算速度提升,功耗降为1/139,计算延迟降为1/26。     

Input Resistance Calculation for Circular Microstrip Antennas Using Adaptive Neuro-Fuzzy Inference System

This paper presents a new method based on adaptive neuro-fuzzy inference system (ANFIS) to calculate the input resistance of circular microstrip patch antennas. The ANFIS is a fuzzy inference system (FIS) implemented in the framework of an adaptive fuzzy neural network. It combines the explicit knowledge representation of FIS with learning power of neural networks. A hybrid learning algorithm based on the least square approach and the backpropagation algorithm is used to optimize the parameters of ANFIS. The input resistance results predicted by ANFIS are in excellent agreement with the experimental results reported elsewhere.     

改进的GMDH型神经网络及其在混沌预测中的应用

本文提出基于改进自组织方法的GMDH（Group Method of Data Handling）型神经网络并将它应用于混沌顶测。一般的GMDH型神经网络的自组织功能是通过给定一个准则阈值来确定或直接给定数值来实现，但GMDH型神经网络的自组织准则的阈值难以合适确定，由此提出了一种简单的自组织方法来实现真正意义上的自组织功能。这种用改进了自组织方法所构成的GMDH型神经网络可以应用于混沌时间序列预测。通过仿真实验，证明其预测效果明显比基本的GMDH型神经网络好，即改进GMDH型神经网络优于基本的GMDH型神经网络。     

Design of Thermally Robust Clock Trees Using Dynamically Adaptive Clock Buffers

On-chip temperature gradient has emerged as a major design concern for high-performance integrated circuits for the current and future technology nodes. Clock skew is an undesirable phenomenon for synchronous digital circuits that is exacerbated by the temperature difference between various parts of the clock tree. The main aim of this paper is to provide intelligent solution for minimizing the temperature-dependent clock skew by designing dynamically adaptive circuit elements, particularly the clock buffers. Using an RLC model of the clock tree, we investigate the effect of on-chip temperature gradient on the clock skew for a number of temperature profiles that can arise in practice due to different architectures and applications. As an effective way of mitigating the variable clock skew, we present an adaptive circuit technique that senses the temperature of different parts of the clock tree and adjusts the driving strengths of the corresponding clock buffers dynamically to reduce the clock skew. Simulation results demonstrate that our adaptive technique is capable of reducing the skew by up to 92.4%, leading to much improved clock synchronization and design performance.      

基于遗传算法的时间序列中频繁结构模式发现研究

本文提出了一个基于小生境遗传算法和模式缓存的时间序列中频繁结构模式的发现算法，该算法具有轻便、灵活、可扩放性好的特点，可根据实际情况合理配置计算时间和所占用的内存资源，并可实现挖掘结果的实时动态更新输出，在实际时间序列数据上的实验证明了该算法的有效性。