The quantum chaos is related to a Gaussian random matrix model, which shows dip-ramp-plateau behavior in the spectral form factor for large size $N$. of time dependent model also with rounding instead kink near Heisenberg time. This converted two made $M_1$ and $M_2$. numerical evaluation finite $N$ analytic expression are compared factor.

In this study, we propose a projection estimation method for large-dimensional matrix factor models with cross-sectionally spiked eigenvalues. By projecting the observation onto row or column space, simplify analysis series to that of lower-dimensional tensor. This also reduces magnitudes idiosyncratic error components, thereby increasing signal-to-noise ratio, because linearly filters matrix. ...