An Introduction to Wavelets and Other Filtering Methods in Finance and EconomicsElsevier, 2001 M10 12 - 359 páginas An Introduction to Wavelets and Other Filtering Methods in Finance and Economics presents a unified view of filtering techniques with a special focus on wavelet analysis in finance and economics. It emphasizes the methods and explanations of the theory that underlies them. It also concentrates on exactly what wavelet analysis (and filtering methods in general) can reveal about a time series. It offers testing issues which can be performed with wavelets in conjunction with the multi-resolution analysis. The descriptive focus of the book avoids proofs and provides easy access to a wide spectrum of parametric and nonparametric filtering methods. Examples and empirical applications will show readers the capabilities, advantages, and disadvantages of each method.
|
Dentro del libro
Resultados 6-10 de 28
Página 32
... weights of the components with 12 and 20 period oscillations are much higher than any other component of the signal. where i = V-1, f is the frequency defined earlier and we is the impulse response function of a filter. Notice that the ...
... weights of the components with 12 and 20 period oscillations are much higher than any other component of the signal. where i = V-1, f is the frequency defined earlier and we is the impulse response function of a filter. Notice that the ...
Página 38
... weights is given by 1 ei2RfI(N–M)/2 sin [2t f(M + N + 1)/2] - ——, (2.24) N + M + 1 sin(27tf/2) H(f) = where M is the number of lagged values and N is the number of future values of the input (Oppenheim and Schafer, 1989, Ch. 6). Figure ...
... weights is given by 1 ei2RfI(N–M)/2 sin [2t f(M + N + 1)/2] - ——, (2.24) N + M + 1 sin(27tf/2) H(f) = where M is the number of lagged values and N is the number of future values of the input (Oppenheim and Schafer, 1989, Ch. 6). Figure ...
Página 43
... weight is given to the current observation. Equation 2.32 may also be modified as 6 = x(r. — F)* + (1–3)(6 1 + st-1), (2.33) where st-1 is obtained from the following equation: s = y(6 – 6: 1) + (1 – y)st-1, (2.34) and 0 < y < 1 ...
... weight is given to the current observation. Equation 2.32 may also be modified as 6 = x(r. — F)* + (1–3)(6 1 + st-1), (2.33) where st-1 is obtained from the following equation: s = y(6 – 6: 1) + (1 – y)st-1, (2.34) and 0 < y < 1 ...
Página 46
... weights; that is, R X = XL wiy—i. (2.37) i=–K The coefficients of the BK filter are derived under the constraint that the filter gain should be zero at zero frequency. This constraint leads to the requirement that the sum of the filter ...
... weights; that is, R X = XL wiy—i. (2.37) i=–K The coefficients of the BK filter are derived under the constraint that the filter gain should be zero at zero frequency. This constraint leads to the requirement that the sum of the filter ...
Página 52
... weights in Equation 3.3 are equal to each other as in a simple moving average and each weight is defined to be w = 1/N so that N-1 1 'E' *N = X wyN- = N XLyw-i, (3.4) i=0 i=0 which is the arithmetic average or mean of the observations ...
... weights in Equation 3.3 are equal to each other as in a simple moving average and each weight is defined to be w = 1/N so that N-1 1 'E' *N = X wyN- = N XLyw-i, (3.4) i=0 i=0 which is the arithmetic average or mean of the observations ...
Contenido
1 | |
15 | |
51 | |
CHAPTER 4 DISCRETE WAVELET TRANSFORMS | 96 |
CHAPTER 5 WAVELETS AND STATIONARY PROCESSES | 161 |
CHAPTER 6 WAVELET DENOISING | 202 |
CHAPTER 7 WAVELETS FOR VARIANCECOVARIANCE ESTIMATION | 235 |
CHAPTER 8 ARTIFICIAL NEURAL NETWORKS | 272 |
NOTATIONS | 315 |
BIBLIOGRAPHY | 323 |
INDEX | 349 |
Otras ediciones - Ver todas
An Introduction to Wavelets and Other Filtering Methods in Finance and Economics Ramazan Gençay,Faruk Selçuk,Brandon Whitcher Sin vista previa disponible - 2002 |
Términos y frases comunes
analysis applied approximate associated assumed basis calculated components computed correlation covariance cycle decomposition defined determined difference discrete distribution dynamics Equation error estimator example exchange feedforward network Figure Fourier transform frequency function gain function Gaussian given Haar hidden units increases indicate input interval known lags length linear matrix mean method MODWT moving average network model neural network noise observations obtained original output parameter performance period phase plotted points prediction presented procedure produce properties random recurrent respectively response returns rule sample scale seasonal sequence shift shows signal simple simulation smooth spectral spectrum squared standard stationary statistical studied term thresholding transform values variables variance vector volatility wavelet coefficients wavelet details wavelet filter wavelet scale wavelet transform wavelet variance weights zero