Given their recent popularity and clear evidence of wide applicability the most of the space in this chapter is devoted to Wavelet transforms. Statistical multiscale modeling has, in recent decade, become a well established area in both theoretical and applied statistics, with impact to developments in statistical methodology.
Wavelet-based methods are important in statistics in areas such as regression, density and function estimation, factor analysis, modeling and forecasting in time series analysis, in assessing self-similarity and fractality in data, in spatial statistics.
The attention of the statistical community was attracted in late 1980's and early 1990's, when Donoho, Johnstone, and their coauthors demonstrated that wavelet thresholding, a simple denoising procedure, had desirable statistical optimality properties. Since then, wavelets have proved useful in many statistical disciplines, notably in nonparametric statistics and time series analysis. Bayesian concepts and modeling approaches have, more recently, been identified as providing promising contexts for wavelet-based denoising applications.
In addition to replacing traditional orthonormal bases in a variety statistical problems, wavelets brought novel techniques and invigorated some of the existing ones.
We start first with a statistical application of wavelet transforms. This example emphasizes specificity of wavelet-based denoising not shared by standard state-of-art denoising techniques.
A researcher in geology was interested in predicting earthquakes by the level of water in nearby wells. She had a large ( measurements) data set of water levels taken every hour in a period of time of about one year in a California well. Here is the description of the problem.
The ability of water wells to act as strain meters has been observed for centuries. The Chinese, for example, have records of water flowing from wells prior to earthquakes. Lab studies indicate that a seismic slip occurs along a fault prior to rupture. Recent work has attempted to quantify this response, in an effort to use water wells as sensitive indicators of volumetric strain. If this is possible, water wells could aid in earthquake prediction by sensing precursory earthquake strain.
We have water level records from six wells in southern California, collected over a six year time span. At least moderate size earthquakes (magnitude -) occurred in close proximity to the wells during this time interval. There is a significant amount of noise in the water level record which must first be filtered out. Environmental factors such as earth tides and atmospheric pressure create noise with frequencies ranging from seasonal to semidiurnal. The amount of rainfall also affects the water level, as do surface loading, pumping, recharge (such as an increase in water level due to irrigation), and sonic booms, to name a few. Once the noise is subtracted from the signal, the record can be analyzed for changes in water level, either an increase or a decrease depending upon whether the aquifer is experiencing a tensile or compressional volume strain, just prior to an earthquake.
A plot of the raw data for hourly measurements over one year ( observations) is given in Fig. 7.6a, with a close-up in Fig. 7.6b. After applying the wavelet transform and further processing the wavelet coefficients (thresholding), we obtained a fairly clean signal with a big jump at the earthquake time. The wavelet-denoised data are given in Fig. 7.6d. The magnitude of the water level change at the earthquake time did not get distorted in contrast to traditional smoothing techniques. This local adaptivity is a desirable feature of wavelet methods.
For example, Fig. 7.6c, is denoised signal after applying supsmo smoothing procedure. Note that the earthquake jump is smoothed, as well.
(a)
(b)
(c) (d) |
The first theoretical results in wavelets had been concerned with continuous wavelet decompositions of functions and go back to the early 1980s. Papers of Morlet et al. (1982)[19] and Grossmann and Morlet (1984, 1985)[13,14] were among the first on this subject.
Let , be a family of functions defined as translations and re-scales of a single function ,
Normalization constant ensures that the norm is independent of and The function (called the wavelet function) is assumed to satisfy the admissibility condition,
Wavelet functions are usually normalized to ''have unit energy'', i.e., .
For example, the second derivative of the Gaussian function,
For any square integrable function , the continuous wavelet transform is defined as a function of two variables
Figure 7.8 gives the doppler test function, , , and its continuous wavelet transform. The wavelet used was Mexican Hat. Notice the distribution of ''energy'' in the time/frequency plane in Fig. 7.8b.
(a)
(b) |
When the admissibility condition is satisfied, i.e., , it is possible to find the inverse continuous transform via the relation known as resolution of identity or Calderón's reproducing identity,
The continuous wavelet transform of a function of one variable is a function of two variables. Clearly, the transform is redundant. To ''minimize'' the transform one can select discrete values of and and still have a lossless transform. This is achieved by so called critical sampling.
The critical sampling defined by
Fundamental for construction of critically sampled orthogonal wavelets is a notion of multiresolution analysis introduced by Mallat (1989a, 1989b)[16,17]. A multiresolution analysis (MRA) is a sequence of closed subspaces in such that they lie in a containment hierarchy
The coefficients in (7.21) are important in efficient application of wavelet transforms. The (possibly infinite) vector , will be called a wavelet filter. It is a low-pass (averaging) filter as will become clear later by its analysis in the Fourier domain.
To further explore properties of multiresolution analysis subspaces and their bases, we will often work in the Fourier domain.
It will be convenient to use Fourier domain for subsequent analysis of wavelet paradigm. Define the function as follows:
In the Fourier domain the relation (7.21) becomes
By iterating (7.23), one gets
Next, we prove two important properties of wavelet filters associated with an orthogonal multiresolution analysis, normalization and orthogonality.
This result also follows from .
For any
An important special case is for which (7.26) becomes
The fact that the system , constitutes an orthonormal basis for can be expressed in the Fourier domain in terms of either or .
In terms of ,
Orthogonality condition 7.29 can be expressed in terms of as:
Since , then by (7.23)
Now split the sum in (7.33) into two sums - one with odd and the other with even indices, i.e.,
Whenever a sequence of subspaces satisfies MRA properties, there exists (though not unique) an orthonormal basis for ,
Next, we discuss the derivation of a wavelet function from the scaling function. Since (because of the containment ), it can be represented as
Define
The spaces and are orthogonal by construction. Therefore,
By taking into account the definitions of and , and by the derivation as in (7.32), we find
From (7.38), we conclude that there exists a function such that
To summarize, we choose such that
Standard choices for are , , and ; however, any other function satisfying (i)-(iii) will generate a valid . We choose to define as
The form of and (7.29) imply that , is an orthonormal basis for .
Since , the orthogonality condition (7.32) can be rewritten as
In addition to their simplicity and formidable applicability, Haar wavelets have tremendous educational value. Here we illustrate some of the relations discussed in the Sect. 7.3.3 using the Haar wavelet. We start with scaling function and pretend that everything else is unknown. By inspection of simple graphs of two scaled Haar wavelets and stuck to each other, we conclude that the scaling equation (7.21) is
The transfer functions are
Relation (7.37) becomes
Although the Haar wavelets are well localized in the time domain, in the frequency domain they decay at the slow rate of and are not effective in approximating smooth functions.
The most important family of wavelets was discovered by Ingrid Daubechies and fully described in Daubechies (1992)[8]. This family is compactly supported with various degrees of smoothness.
The formal derivation of Daubechies' wavelets goes beyond the scope of this chapter, but the filter coefficients of some of its family members can be found by following considerations.
For example, to derive the filter taps of a wavelet with vanishing moments, or equivalently, filter taps, we use the following equations.
The normalization property of scaling function implies
We obtained equations with unknowns; however the system is solvable since the equations are not linearly independent.
For , the system is
Figure 7.9 depicts two scaling function and wavelet pairs from the Daubechies family. Figure 7.9a,b depict the pair with two vanishing moments, while Fig. 7.9c,d depict the pair with four vanishing moments.
(a)
(b)
(c) (d) |