auto associative memory

Computing paradigm based on von Neuman architectures cannot keep up with the ever-increasing data growth (also called “data deluge gap”). This has resulted in investigating novel computing paradigms and design approaches at all levels from materials to system-level implementations and applications. An alternative computing approach based on artificial neural networks uses oscillators to compute or Oscillatory Neural Networks (ONNs). ONNs can perform computations efficiently and can be used to build a more extensive neuromorphic system. Here, we address a fundamental problem: can we efficiently perform artificial intelligence applications with ONNs? We present a digital ONN implementation to show a proof-of-concept of the ONN approach of “computing-in-phase” for pattern recognition applications. To the best of our knowledge, this is the first attempt to implement an FPGA-based fully-digital ONN. We report ONN accuracy, training, inference, memory capacity, operating frequency, hardwa...

Vector LIDA was proposed in 2014 as major overhaul to the representation system in the LIDA model. It replaced the nodes and links representation system used earlier with Modular Composite Representation (Modular Composite Representation) vectors for representing knowledge in LIDA. Although, it was mostly trivial to swap out nodes and links with these vectors, as described in the Vector LIDA paper, it was not trivial to obtain these vectors from sensors so that symbol grounding is maintained. In this paper, we propose a sensory memory system and a method for obtaining Modular Composite Representation vectors from it in a way that enables symbol grounding. We propose to build sensory memory out of Hierarchical Temporal Memory’s Cortical Learning Algorithm’s regions, an online learning algorithm, which produce sparse distributed representations of spatiotemporal patterns recognized from its stimulus from the environment. These sparse distributed representations can then be translated ...

In the present paper, an effort has been made to compare and analyze the performance for pattern recalling with conventional hebbian learning rule and with evolutionary algorithm in Hopfield Model of feedback Neural Networks. A set of ten objects has been considered as the pattern set. In the Hopfield type of neural networks of associative memory, the weighted code of input patterns provides an auto-associative function in the network. The storing of the objects has been performed using Hebbian rule and recalling of these stored patterns on presentation of prototype input patterns has been made using both-conventional hebbian rule and evolutionary algorithm. Exploration of the population generation techniques (mutation and elitism), crossover and setting up of proper fitness evaluation functions to generate the new population of the weight matrices from the optimal weight matrix of the stored patterns has been done. The simulated results show that the genetic algorithm is the best searching technique to recall the approximate input patterns.

In the present paper, an effort has been made to compare and analyze the performance for pattern recalling with conventional hebbian learning rule and with evolutionary algorithm in Hopfield Model of feedback Neural Networks. A set of ten objects has been considered as the pattern set. In the Hopfield type of neural networks of associative memory, the weighted code of input patterns provides an auto-associative function in the network. The storing of the objects has been performed using Hebbian rule and recalling of these stored patterns on presentation of prototype input patterns has been made using both-conventional hebbian rule and evolutionary algorithm. Exploration of the population generation techniques (mutation and elitism), crossover and setting up of proper fitness evaluation functions to generate the new population of the weight matrices from the optimal weight matrix of the stored patterns has been done. The simulated results show that the genetic algorithm is the best searching technique to recall the approximate input patterns.

In the present paper, an effort has been made to compare and analyze the performance for pattern recalling with conventional hebbian learning rule and with evolutionary algorithm in Hopfield Model of feedback Neural Networks. A set of ten objects has been considered as the pattern set. In the Hopfield type of neural networks of associative memory, the weighted code of input patterns provides an auto-associative function in the network. The storing of the objects has been performed using Hebbian rule and recalling of these stored patterns on presentation of prototype input patterns has been made using both-conventional hebbian rule and evolutionary algorithm. Exploration of the population generation techniques (mutation and elitism), crossover and setting up of proper fitness evaluation functions to generate the new population of the weight matrices from the optimal weight matrix of the stored patterns has been done. The simulated results show that the genetic algorithm is the best searching technique to recall the approximate input patterns.

In this paper we are studying the performance of Hopfield neural network for recalling of memorized patterns from the Hebbian rule and genetic algorithm for English characters. In this process the genetic algorithm is employed in random form and sub-optimal form for recalling of memorized patterns corresponding to the presented noisy prototype input patterns. The objective of this study is to determine the optimal weight matrix for correct recalling corresponding to noisy form of the English characters. In this study the performance of neural network is evaluated in terms of the rate of success for recalling of noisy input patterns of the English characters with GA in two aspects. The first aspect reflects the random nature of the GA and the second one exhibits the suboptimal nature of the GA for its exploration. The simulated results demonstrate the better performance of network for recalling of the stored letters of English alphabets using genetic algorithm on the suboptimal weight matrix.

Page 1. 200 Int. J. Business Information Systems, Vol. 6, No. 2, 2010 Copyright © 2010 Inderscience Enterprises Ltd. Pattern recalling analysis of English alphabets using Hopfield model of feedback neural network with evolutionary searching Somesh Kumar* ...

Hyperdimensional Computing is an emergent model of computation where all objects are represented in high-dimensional vectors. This model includes a well-defined set of arithmetic operations that produce new highdimensional vectors, which, in addition to represent basic entities, can also represent more complex data structures such as sets, relations and sequences. This paper presents a method for sequence prediction using Hyperdimensional Computing and the Sparse Distributed Memory model. The proposed method is based on the encoding, storage and retrieval of sequence vectors, which store the k consecutive vectors of a sequence. The next element of a sequence is selected by taking into account the current, as well as the k immediate preceding elements of the sequence. Each vector is associated to a sequence vector that is stored in memory; the way in which each vector is associated to its sequence vectors is the main contribution of this paper. We present experimental results for the encoding and prediction of randomly generated sequences and the results indicate that the method performs correct predictions.

In this paper, implementation of a genetic algorithm has been described to store and later, recall of some prototype patterns in Hopfield neural network associative memory. Various operators of genetic algorithm (mutation, cross-over, elitism etc) are used to evolve the population of optimal weight matrices for the purpose of storing the patterns and then recalling of the patterns with induced noise was made, again using a genetic algorithm. The optimal weight matrices obtained during the training are used as seed for starting the GA in recalling, instead starting with random weight matrix. A detailed study of the comparison of results thus obtained with the earlier results has been done. It has been observed that for Hopfield neural networks, recall of patterns is more successful if evolution of weight matrices is applied for training purpose also.

Sparse distributed memory is an auto-associative memory system that stores high dimensional Boolean vectors. Here we present an extension of the original SDM, the Integer SDM that uses modular arithmetic integer vectors rather than binary vectors. This extension preserves many of the desirable properties of the original SDM: auto-associativity, content addressability, distributed storage, and robustness over noisy inputs. In addition, it improves the representation capabilities of the memory and is more robust over normalization. It can also be extended to support forgetting and reliable sequence storage. We performed several simulations that test the noise robustness property and capacity of the memory. Theoretical analyses of the memory's fidelity and capacity are also presented. (J. Snaider).

The combination of evolutionary algorithms and ANN has been a recent interest in the field of research. Hopfield model is a type of recurrent neural network which has been widely studied for the purpose of associative memories. In the present work, this Hopfield Model of feedback neural networks has been studied with Monte Carlo adaptation learning rule and one evolutionary searching algorithm i.e.