Figure 6

Comparisons between these two methods (i.e. QR-CS and CS methods) on the reconstruction success rates as function nt changes in Celegansneural, Dolphin, Football, ZK, Jazz and Polbooks networks. The values of networks size N and their average sparsities are shown in Table 1. These experiments select the measurement data from the time \(t=350\), nt is the ratio between the row and the column of the measurement matrix (in the QR-CS method, \(nt=(P-N)/M\), \(N+1\le P\le 2N\), and in the CS method (where state matrix X is replaced by stochastic Gaussian matrix), \(nt=P/N\), \(1\le P\le N\)), and \(M=N\). The input vector u is the \(M\times P\)-dimensional standard Gaussian noise. The success rate is defined as the ratio between the simulation number of successful reconstruction \(\alpha \) and the simulation number \(\beta \). In these experiments, 20 simulations are performed, and the error of each simulation is \(\varepsilon < {10}^{-6}\).