Fig. 7From: A general deep learning framework for network reconstruction and dynamics learningAccuracy of reconstruction versus the size of training data. The amount of training data is ranging from about 102 to 103 with each sampling evolving for T=100 time steps. The results are obtained on CML model with 10 nodes and the network topology is random 4-regular graphs. We set s=0.2 and r=3.6 to generate chaotic dynamicsBack to article page