Figure 2
From: Optimizing a quantum reservoir computer for time series prediction
(a) Memory capacity shown as a function of the input timescale \(\Delta t\). The values for h and J are sampled from \({\mathscr {U}}[0.5-h_s,0.5+h_s]\) and \({\mathscr {U}}[0.5-J_s,0.5+J_s]\), respectively. Deviation in J leads to better performance but the deviation in h is less important. Furthermore, for all parameters simulated, there exists an optimal timescale for \(\Delta t\), which maximizes the capacity. (b) Readout signals Tr\([Z_i \rho (t)]\) for three different time windows \([t_i,t_i+\Delta t]\) for parameters \(h_s=0.01\), \(J_s=0.5\). The leftmost, center and the rightmost panels corresponds to timescales shorter, equal and longer than the optimal timescale \(\Delta t_{\text {opt}}\), respectively.
