Abstract
This paper introduces a new model of associative memory, capable of both binary and continuous-valued inputs. Based on kernel theory, the memory model is on one hand a generalization of Radial Basis Function networks and, on the other, is in feature space, analogous to a Hopfield network. Attractors can be added, deleted, and updated on-line simply, without harming existing memories, and the number of attractors is independent of input dimension. Input vectors do not have to adhere to a fixed or bounded dimensionality; they can increase and decrease it without relearning previous memories. A memory consolidation process enables the network to generalize concepts and form clusters of input data, which outperforms many unsupervised clustering techniques; this process is demonstrated on handwritten digits from MNIST. Another process, reminiscent of memory reconsolidation is introduced, in which existing memories are refreshed and tuned with new inputs; this process is demonstrated on series of morphed faces.
Highlights
Memory experiments demonstrated persistent activity in several structures in the lower mammal, primate, and human brains including the hippocampus [1,2], prefrontal [3], visual [4] and oculomotor cortex [5], basal ganglia [6], etc
For the first time in neural memory models, we have demonstrated a method allowing input dimension not to be constrained to a fixed size or be a priori bounded; dimension can change with time, similar to organic memory allocation for memories of greater importance or increased detail
2.1 Our Kernel Heteroassociative Memory Model A general framework of heteroassociative memory can be defined on the input Ex and output Ey spaces, with dimensionalities n and p respectively, and with m pairs of vectors xi[Ex, yi[Ey, i~1 . . . m to be stored
Summary
Memory experiments demonstrated persistent activity in several structures in the lower mammal, primate, and human brains including the hippocampus [1,2], prefrontal [3], visual [4] and oculomotor cortex [5], basal ganglia [6], etc. We introduce a memory model, in which its memory attractors do not lie in the input or neural space as in classical models but rather in a feature space with large or infinite dimension. For the first time in neural memory models, we have demonstrated a method allowing input dimension not to be constrained to a fixed size or be a priori bounded; dimension can change with time, similar to organic memory allocation for memories of greater importance or increased detail. These attributes may be very beneficial in psychological modeling
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.