N j is the number of modes in the jth cluster, and W ij is the correlation weight of mode X i and cluster j.
In the formula above, W ij is the correlation weight of mode X i and cluster j, X i (i = 1, ⋯, N) is the position of the ith mode, and Z j (j = 1, ⋯, K) is the center of the jth cluster, which can be found in Eq. (2): {Z}_j=frac{1}{N_j}{displaystyle sum_{i=1}^N{w}_{ij}{x}_i} (2).
In the case of the correlation weight, the support of the wavelets plays only a minor role, as this weight is by itself computed in a window on the approximation coefficients.
Table 3 Average within-subject inter-method correlation of weights of individual attributes 1.
Several authors (Goeman and Bühlmann, 2007; Gold et al., 2007; Huang et al., 2009) have raised the issue of correlation between weights of entities: generally the weights of biologically related genes or proteins change together and therefore a null model assuming independence between weights may result in exaggerated P-values.
Benefiting from the new formulation of correlation weight joint data vector, the method does not need to calculate the noise subspace dimension before estimating the InSAR interferometric phase, thus the introduced trouble will be avoided.
In this article, we propose a method to estimate the synthetic aperture radar interferometry (InSAR) interferometric phase based on the model of correlation weight joint pixel by using the joint subspace projection technique.
Figure 1 Formulation of the correlation weight joint data vector.
Figure 4 Formulation of the correlation weight data vector when the coregistration error is unknown.
β is a positive integer and serves to transform the correlations so that weights of large correlation differences are given more emphasis compared to smaller differences which are less meaningful.
