5,337
edits
Deep Learning: Difference between revisions
→Strongly local min-max
| Line 1,310: | Line 1,310: | ||
\end{aligned} | \end{aligned} | ||
</math> | </math> | ||
For asymptotic stability, we need <math>| \lambda_{\max} (J) | < 1</math>. | |||
<math>J = I + \eta H</math> with lr <math>\eta > 0</math> | |||
;Lemma | |||
If H has eigenvalue with negative real parts then eigenvalues of J lie in a unit ball iff <math>\eta < \frac{1}{|Re(\lambda)} \frac{2}{1+ (Im(\lambda)/Re(\lambda))^2}</math> for all <math>\lambda</math>. | |||
==Misc== | ==Misc== | ||