Inverse Ising techniques to infer underlying mechanisms from data
Zeng Hong-Li1, 2, †, Aurell Erik3, 4, ‡
       

Scatter-plots of inferred epistatic fitness against the true fitness based on the averaged results from singletime: (a) with sad face: r = 0.1, where the KNS theory cannot be satisfied here, (b) with smiling face: r = 0.5, where inference works, (c) with not-that-sad face: r = 0.9, where the KNS theory works but with very heavy fluctuations. Here the DCA-nMF algorithm for Jij’s is utilized.