Skip to main content
Fig. 16 | Applied Network Science

Fig. 16

From: Evolving network representation learning based on random walks

Fig. 16

EVONRL scalability (running time axis is in logarithmic scale). StaticNRL scales linearly to the number of new edges added in the network, since it has to run again and again for every new edge. At the same time, EVONRL is able to accommodate the changes more than 100 times faster than StaticNRL. This behavior is even more stressed in the larger network (where the number of nodes is larger)

Back to article page