On the biological plausibility of orthogonal initialisation for solving gradient instability in deep neural networks

10/27/2022
by   Nikolay Manchev, et al.
0

Initialising the synaptic weights of artificial neural networks (ANNs) with orthogonal matrices is known to alleviate vanishing and exploding gradient problems. A major objection against such initialisation schemes is that they are deemed biologically implausible as they mandate factorization techniques that are difficult to attribute to a neurobiological process. This paper presents two initialisation schemes that allow a network to naturally evolve its weights to form orthogonal matrices, provides theoretical analysis that pre-training orthogonalisation always converges, and empirically confirms that the proposed schemes outperform randomly initialised recurrent and feedforward networks.

READ FULL TEXT

Please sign up or login with your details

Forgot password? Click here to reset