Author response for "Multi‐manifold NIRS modelling via stacked contractive auto‐encoders"