Distributed gradient descent in continuous time almost surely converges to local minima, because saddle points only attract initializations from a lower-dimensional stable manifold.
An overview of recent pr ogress in the study of distributed multi-agent coordination,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
math.OC 1years
2019 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Distributed Gradient Descent: Nonconvergence to Saddle Points and the Stable-Manifold Theorem
Distributed gradient descent in continuous time almost surely converges to local minima, because saddle points only attract initializations from a lower-dimensional stable manifold.