Fig. 9

Results of running NOL* algorithms on the same networks as Fig. 3, but starting from initial samples using random walk with jump sampling rather than node sampling with induction. Results are similar to those presented in the main text. In general, the performance of NOL is more consistent on random walk samples, while the performance of NOL-HTR is approximately the same or slightly worse (e.g. BTER). Interestingly, the low degree heuristic also appears to perform better on random walk samples