Direct Feedback alignment provides learning in deep neural nets
 from {1423}
 Feedback alignment is able to provide zero training error even in convolutional networks and very deep networks, completely without error backpropagation.
 Biologically plausible: error signal is entirely local, no symmetric or reciprocal weights required.
 Still, it requires supervision.
 Almost as good as backprop!

 Clearly written, easy to follow math.
 Though the proof that feedbackalignment direction is within 90 deg of backprop is a bit impenetrable, needs some reorganization or additional exposition / annotation.
 3x400 tanh network tested on MNIST; performs similarly to backprop, if faster.
 Also able to train very deep networks, on MNIST  CIFAR10, CIFAR100, 100 layers (which actually hurts this task).
