Figure 4

Network reconstruction performances. (a,c,e,g) Element values \({B}_{ij}\) vs. different nt for the linear network system. (b,d,f,h) Relationship between reconstruction success rate and nt of network structure matrix A and matrix B for four model networks (i.e. ER, NW, BA and WS networks). The networks size N is 100 with average node sparsity \(\langle k\rangle =4\). The elements of matrix B are randomly selected as 0 or 1, and the sparsity of controlled matrix B is 4 (namely \({\parallel {\bar{B}}^{T}\parallel }_{0}=4\)). The input vector u is the \(M\times P\)-dimensional standard Gaussian noise. These experiments select measurement data from the time \(t=350\), nt is the ratio between the row and the column of the measurement matrix (\(nt=(P-N)/M\)), P is the number of experiments \((N+1\le P\le 3N)\), and \(M=200\). The success rate is defined as the ratio between the number of successful simulation \(\alpha \) and the simulation number \(\beta \). In these experiments, 20 simulations were performed, and the error of each simulation is \(\varepsilon < {10}^{-6}\).