Graph isomorphism in quasipolynomial time I: The “Local Certificates Algorithm.” Combinatorics and Theoretical Computer Science seminar, Chicago, November 10, 2015.

As with individual perceptrons, multi-layer networks can be used for learning tasks. However, the learning algorithm that we look at (the backpropagation routine) is derived mathematically, using ...

The Forward-Forward algorithm (FF) is comparable in speed to backpropagation but has the advantage that it can be used when the precise details of the forward computation are unknown.

Backpropagation is not limited to function derivatives. Any algorithm that effectively takes the loss function and applies gradual, positive changes back through the network is valid.

"Professor tackles graph mining challenges with new algorithm." ScienceDaily. ScienceDaily, 18 October 2024. <www.sciencedaily.com / releases / 2024 / 10 / 241018162554.htm>.

Since the 1970s, algorithms have been able to test graph isomorphism, but in exponential time. This means that the increasing complexity of the graphs increased the algorithm's running time ...

Neural networks using the backpropagation algorithm were biologically “unrealistic in almost every respect” he said. For one thing, neurons mostly send information in one direction.