Figure 2

The aggregating curves of dynamic network filtering methods. The aggregating curves for fastviz fastviz (black line), rectangular sliding time-window (green dashed), and exponential time-window (blue dotted). The steps of the fastviz fastviz method correspond to consecutive multiplications by the forgetting factor $C_{\mathrm{f}}=2/3$ performed after each forgetting period $T_{\mathrm{f}}$. The rectangular time-window width is set to $T_{\mathrm{tw}}=3T_{\mathrm{f}}$. The exponent of the exponentially decaying time-window corresponds to the forgetting factor of fastvizfastviz. For these values of the parameters, areas under the aggregating curves of both methods are approximately equal, according to Equation 1.