• 제목/요약/키워드: decomposition series

검색결과 252건 처리시간 0.027초

Wavelet Thresholding Techniques to Support Multi-Scale Decomposition for Financial Forecasting Systems

  • Shin, Taeksoo;Han, Ingoo
    • 한국데이타베이스학회:학술대회논문집
    • /
    • 한국데이타베이스학회 1999년도 춘계공동학술대회: 지식경영과 지식공학
    • /
    • pp.175-186
    • /
    • 1999
  • Detecting the features of significant patterns from their own historical data is so much crucial to good performance specially in time-series forecasting. Recently, a new data filtering method (or multi-scale decomposition) such as wavelet analysis is considered more useful for handling the time-series that contain strong quasi-cyclical components than other methods. The reason is that wavelet analysis theoretically makes much better local information according to different time intervals from the filtered data. Wavelets can process information effectively at different scales. This implies inherent support fer multiresolution analysis, which correlates with time series that exhibit self-similar behavior across different time scales. The specific local properties of wavelets can for example be particularly useful to describe signals with sharp spiky, discontinuous or fractal structure in financial markets based on chaos theory and also allows the removal of noise-dependent high frequencies, while conserving the signal bearing high frequency terms of the signal. To date, the existing studies related to wavelet analysis are increasingly being applied to many different fields. In this study, we focus on several wavelet thresholding criteria or techniques to support multi-signal decomposition methods for financial time series forecasting and apply to forecast Korean Won / U.S. Dollar currency market as a case study. One of the most important problems that has to be solved with the application of the filtering is the correct choice of the filter types and the filter parameters. If the threshold is too small or too large then the wavelet shrinkage estimator will tend to overfit or underfit the data. It is often selected arbitrarily or by adopting a certain theoretical or statistical criteria. Recently, new and versatile techniques have been introduced related to that problem. Our study is to analyze thresholding or filtering methods based on wavelet analysis that use multi-signal decomposition algorithms within the neural network architectures specially in complex financial markets. Secondly, through the comparison with different filtering techniques' results we introduce the present different filtering criteria of wavelet analysis to support the neural network learning optimization and analyze the critical issues related to the optimal filter design problems in wavelet analysis. That is, those issues include finding the optimal filter parameter to extract significant input features for the forecasting model. Finally, from existing theory or experimental viewpoint concerning the criteria of wavelets thresholding parameters we propose the design of the optimal wavelet for representing a given signal useful in forecasting models, specially a well known neural network models.

  • PDF

Wavelet Thresholding Techniques to Support Multi-Scale Decomposition for Financial Forecasting Systems

  • Shin, Taek-Soo;Han, In-Goo
    • 한국지능정보시스템학회:학술대회논문집
    • /
    • 한국지능정보시스템학회 1999년도 춘계공동학술대회-지식경영과 지식공학
    • /
    • pp.175-186
    • /
    • 1999
  • Detecting the features of significant patterns from their own historical data is so much crucial to good performance specially in time-series forecasting. Recently, a new data filtering method (or multi-scale decomposition) such as wavelet analysis is considered more useful for handling the time-series that contain strong quasi-cyclical components than other methods. The reason is that wavelet analysis theoretically makes much better local information according to different time intervals from the filtered data. Wavelets can process information effectively at different scales. This implies inherent support for multiresolution analysis, which correlates with time series that exhibit self-similar behavior across different time scales. The specific local properties of wavelets can for example be particularly useful to describe signals with sharp spiky, discontinuous or fractal structure in financial markets based on chaos theory and also allows the removal of noise-dependent high frequencies, while conserving the signal bearing high frequency terms of the signal. To data, the existing studies related to wavelet analysis are increasingly being applied to many different fields. In this study, we focus on several wavelet thresholding criteria or techniques to support multi-signal decomposition methods for financial time series forecasting and apply to forecast Korean Won / U.S. Dollar currency market as a case study. One of the most important problems that has to be solved with the application of the filtering is the correct choice of the filter types and the filter parameters. If the threshold is too small or too large then the wavelet shrinkage estimator will tend to overfit or underfit the data. It is often selected arbitrarily or by adopting a certain theoretical or statistical criteria. Recently, new and versatile techniques have been introduced related to that problem. Our study is to analyze thresholding or filtering methods based on wavelet analysis that use multi-signal decomposition algorithms within the neural network architectures specially in complex financial markets. Secondly, through the comparison with different filtering techniques results we introduce the present different filtering criteria of wavelet analysis to support the neural network learning optimization and analyze the critical issues related to the optimal filter design problems in wavelet analysis. That is, those issues include finding the optimal filter parameter to extract significant input features for the forecasting model. Finally, from existing theory or experimental viewpoint concerning the criteria of wavelets thresholding parameters we propose the design of the optimal wavelet for representing a given signal useful in forecasting models, specially a well known neural network models.

  • PDF

철도차량용 폐 복합소재에서의 탄소섬유 회수 (The Recovery of Carbon Fiber from Carbon Fiber Reinforced Epoxy Composites for Train Body)

  • 이석호;이철규;김용기;김정석;주창식
    • 한국철도학회:학술대회논문집
    • /
    • 한국철도학회 2008년도 추계학술대회 논문집
    • /
    • pp.406-415
    • /
    • 2008
  • Recently, the amount of thermosetting plastic wastes have increased with the production of reinforced plastic composites and causes serious environmental problems. The epoxy composites, one of the versatile thermosetting plastics with excellent properties, cannot be melted down and remolded as what is done in the thermoplastic industry. In this research, a series of experiments that recovers carbon fibers from carbon fiber reinforced epoxy composites for train body was performed. We experimentally examined various decomposition processes and compared their decomposition efficiencies and mechanical property of recovered carbon fibers. For the prevention of tangle of recovered carbon fibers, each composites specimen was fixed with a Teflon supporter and no mechanical mixing was applied. Decomposition products were analyzed by scanning electron microscope (SEM), gas chromatography mass spectrometer (GC-MS), and universal testing machine (UTM). Carbon fibers could be completely recovered from decomposition process using nitric acid aqueous solution, liquid-phase thermal cracking and pyrolysis. The tensile strength losses of the recovered carbon fibers were less than 4%.

  • PDF

전자빔 용접에서 SVD을 이용한 온라인 모니터링 (On-line Monitoring Using SVD in a Electron Beam Welding)

    • Journal of Welding and Joining
    • /
    • 제18권1호
    • /
    • pp.97-103
    • /
    • 2000
  • Time series analysis results show the SVD is a candidate of on-line monitoring of welding penetration when the covariance matrix of a full penetration is used as a mapping function. As the reconstructed embedding vectors from the chaotic scalar time series are manipulated by the covariance matrix, the mapped tim series lie on a hyper-ellipsoid which the lengths of semi-axes are the squared eigenvalues of the covariance matrix in the case of full penetration. These visualize by two dimensional stroboscope views. The other cases like partial penetration, are different in the sense of sizes and shapes. Here we test two types of time series; the ion current and the X-ray. The ion current is better than the X-ray as an on-line monitoring signal, because the difference of the eigenvalue spectrum of the ion(between the pull penetration and partial penetration) is bigger than those of the X-ray.

  • PDF

CERTAIN RADIALLY DILATED CONVOLUTION AND ITS APPLICATION

  • Rhee, Jung-Soo
    • 호남수학학술지
    • /
    • 제32권1호
    • /
    • pp.101-112
    • /
    • 2010
  • Using some interesting convolution, we find kernels recovering the given function f. By a slight change of this convolution, we obtain an identity filter related to the Fourier series in the discrete time domain. We also introduce some techniques to decompose an impulse into several dilated pieces in the discrete domain. The detail examples deal with specific constructions of those decompositions. Also we obtain localized moving averages from a decomposition of an impulse to make hybrid Bollinger bands, that might give various strategies for stock traders.

A FAMILY OF SERIES AND INTEGRALS INVOLVING WHITTAKER, BESSEL FUNCTIONS, AND THEIR PRODUCTS DERIVABLE FROM THE REPRESENTATION OF THE GROUP SO(2, 1)

  • Choi, Junesang;Shilin, I.A.
    • 대한수학회논문집
    • /
    • 제32권4호
    • /
    • pp.999-1008
    • /
    • 2017
  • By mainly using certain properties arising from the semisimple Lie group SO(2, 1), we aim to show how a family of some interesting formulas for bilateral series and integrals involving Whittaker, Bessel functions, and their product can be obtained.

Structural modal identification through ensemble empirical modal decomposition

  • Zhang, J.;Yan, R.Q.;Yang, C.Q.
    • Smart Structures and Systems
    • /
    • 제11권1호
    • /
    • pp.123-134
    • /
    • 2013
  • Identifying structural modal parameters, especially those modes within high frequency range, from ambient data is still a challenging problem due to various kinds of uncertainty involved in vibration measurements. A procedure applying an ensemble empirical mode decomposition (EEMD) method is proposed for accurate and robust structural modal identification. In the proposed method, the EEMD process is first implemented to decompose the original ambient data to a set of intrinsic mode functions (IMFs), which are zero-mean time series with energy in narrow frequency bands. Subsequently, a Sub-PolyMAX method is performed in narrow frequency bands by using IMFs as primary data for structural modal identification. The merit of the proposed method is that it performs structural identification in narrow frequency bands (take IMFs as primary data), unlike the traditional method in the whole frequency space (take original measurements as primary data), thus it produces more accurate identification results. A numerical example and a multiple-span continuous steel bridge have been investigated to verify the effectiveness of the proposed method.

An integrated system for synthesis of plant-wide control structure

  • Choi, In-Seok;Yoon, En-Sup
    • 제어로봇시스템학회:학술대회논문집
    • /
    • 제어로봇시스템학회 1990년도 한국자동제어학술회의논문집(국제학술편); KOEX, Seoul; 26-27 Oct. 1990
    • /
    • pp.1265-1270
    • /
    • 1990
  • A prototype integrated system and its theories for distributed SISO control structure synthesis of complete chemical plants is developed. The scope of this work includes control structure synthesis not only of simple units with unspecified control loops but also of the complex process at preliminary and basic design stage. Hierarchical approach and dual-decomposition strategy (that is multi-layer decomposition and multi-echelon decomposition) is applied to this system. Because automatic control structure synthesis of complex plants is a problem defined as a series of knowledge-intensive tasks within multiple spaces, the established methodology is complemented by not only techniques from knowledge-based expert systems but also shortcut and rigorous control theories. This system is used for education of control designers, process engineers, operators and students as well as for operability studying, in-line and on-line process control structure synthesis.

  • PDF

Modal transformation tools in structural dynamics and wind engineering

  • Solari, Giovanni;Carassale, Luigi
    • Wind and Structures
    • /
    • 제3권4호
    • /
    • pp.221-241
    • /
    • 2000
  • Structural dynamics usually applies modal transformation rules aimed at de-coupling and/or minimizing the equations of motion. Proper orthogonal decomposition provides mathematical and conceptual tools to define suitable transformed spaces where a multi-variate and/or multi-dimensional random process is represented as a linear combination of one-variate and one-dimensional uncorrelated processes. Double modal transformation is the joint application of modal analysis and proper orthogonal decomposition applied to the loading process. By adopting this method the structural response is expressed as a double series expansion in which structural and loading mode contributions are superimposed. The simultaneous use of the structural modal truncation, the loading modal truncation and the cross-modal orthogonality property leads to efficient solutions that take into account only a few structural and loading modes. In addition the physical mechanisms of the dynamic response are clarified and interpreted.

Proper orthogonal decomposition in wind engineering - Part 1: A state-of-the-art and some prospects

  • Solari, Giovanni;Carassale, Luigi;Tubino, Federica
    • Wind and Structures
    • /
    • 제10권2호
    • /
    • pp.153-176
    • /
    • 2007
  • The Proper Orthogonal Decomposition (POD) is a statistical method particularly suitable and versatile for dealing with many problems concerning wind engineering and several other scientific and humanist fields. POD represents a random process as a linear combination of deterministic functions, the POD modes, modulated by uncorrelated random coefficients, the principal components. It owes its popularity to the property that only few terms of the series are usually needed to capture the most energetic coherent structures of the process, and a link often exists between each dominant mode and the main mechanisms of the phenomenon. For this reason, POD modes are normally used to identify low-dimensional subspaces appropriate for the construction of reduced models. This paper provides a state-of-the-art and some prospects on POD, with special regard to its framework and applications in wind engineering. A wide bibliography is also reported.