A Spiking Neuromorphic Architecture Using Gated-RRAM for Associative Memory

This work reports a spiking neuromorphic architecture for associative memory simulated in a SPICE environment using recently reported gated-RRAM (resistive random-access memory) devices as synapses alongside neurons based on complementary metal-oxide semiconductors (CMOSs). The network utilizes a Verilog A model to capture the behavior of the gated-RRAM devices within the architecture. The model uses parameters obtained from experimental gated-RRAM devices that were fabricated and tested in this work. Using these devices in tandem with CMOS neuron circuitry, our results indicate that the proposed architecture can learn an association in real time and retrieve the learned association when incomplete information is provided. These results show the promise for gated-RRAM devices for associative memory tasks within a spiking neuromorphic architecture framework.

Cognitive mechanisms of statistical learning and segmentation of continuous sensory input

Two classes of cognitive mechanisms have been proposed to explain segmentation of continuous sensory input into discrete recurrent constituents: clustering and boundary-finding mechanisms. Clustering mechanisms are based on identifying frequently co-occurring elements and merging them together as parts that form a single constituent. Bracketing (or boundary-finding) mechanisms work by identifying rarely co-occurring elements that correspond to the boundaries between discrete constituents. In a series of behavioral experiments, I tested which mechanisms are at play in the visual modality both during segmentation of a continuous syllabic sequence into discrete word-like constituents and during recognition of segmented constituents. Additionally, I explored conscious awareness of the products of statistical learning—whole constituents versus merged clusters of smaller subunits. My results suggest that both online segmentation and offline recognition of extracted constituents rely on detecting frequently co-occurring elements, a process likely based on associative memory. However, people are more aware of having learnt whole tokens than of recurrent composite clusters.

Multisensory integration of contextual boundaries affects temporal order memory, but not encoding or recognition

We memorize our daily life experiences, which are often multisensory in nature, by segmenting them into distinct event models, in accordance with perceived contextual or situational changes. However, very little is known about how multisensory integration affects segmentation, as most studies have focused on unisensory (visual or audio) segmentation. In three experiments, we investigated the effect of multisensory integration on segmentation in memory and perception. In Experiment 1, participants encoded lists of visual objects while audio and visual contexts changed synchronously or asynchronously. After each list, we tested recognition and temporal associative memory for pictures that were encoded in the same audio-visual context or that crossed a synchronous or an asynchronous multisensory change. We found no effect of multisensory integration for recognition memory: Synchronous and asynchronous changes similarly impaired recognition for pictures encoded at those changes, compared to pictures encoded further away from those changes. Multisensory integration did affect temporal associative memory, which was worse for pictures encoded at synchronous than at asynchronous changes. Follow up experiments showed that this effect was not due to the higher complexity of multisensory over unisensory contexts (Experiment 2), nor that it was due to the temporal unpredictability of contextual changes inherent to Experiment 1 (Experiment 3). We argue that participants formed situational expectations through multisensory integration, such that synchronous multisensory changes deviated more strongly from those expectations than asynchronous changes. We discuss our findings in light of supportive and conflicting findings of uni- and multisensory segmentation.

Pendeteksi Pengenalan Emosi Pada Manusia Menggunakan Hidden Markov Model Dan Bidirectional Associative Memory Dengan Suara

Sistem pengendalian emosi seseorang melalui suara dapat menggunakan algoritma Hidden Markov Model (HMM). Namun, untuk melihat kinerja algoritma HMM dalam sistem aplikasi sudah optimal atau belum, diperlukan suatu perbandingan agar memperoleh hasil yang lebih maksimal. Maka dari itu peneliti melakukan unjuk kerja pengenalian emosi seseorang dengan menggunakan algoritma HMM dan algoritma Bidirectional Associative Memory (BAM) melalui suara. Hidden Markov Model (HMM) terdiri dari rantai markov pada bagian pertama yang menyembunyikan state, oleh karena itu perilaku internal model tetap tidak terlihat. Sedangkan algoritma BAM dapat memproses input yang tidak lengkap, karena adanya hubungan timbal balik antara dari lapisan output ke lapisan input. Pada algoritma BAM, nilai suara pengujian dan nilai sampel suara pelatihan yang diperoleh akan dicari nilai vektornya menggunakan pencarian nilai bobot yang dilakukan dengan cara mengubah matriks biner ke dalam matriks bipolar. Pada penelitian ini akan membuat sebuah sistem aplikasi yang dapat mendeteksi suara dalam bentuk emosi marah, bahagia, dan netral. Dan database yang digunakan adalah suara dari rekaman film. Penelitian ini dilakukan untuk menghasilkan sistem yang dapat mengenali probabilitas emosi pada kategori marah, bahagia dan netral, yaitu dengan menunjukkan unjuk kerja dari kedua metode sehingga kita dapat mengetahui metode mana menghasilkan output yang maksimal.

A Biometric AsymmetricCryptosystem Software Module Based on Convolutional Neural Networks

During the research, the analysis of the existing biometric cryptographic systems was carried out. Some methods that help to generate biometric features were considered and compared with a cryptographic key. For comparing compact vectors of biometric images and cryptographic keys, the following methods are analyzed: designing and training of bidirectional associative memory; designing and training of single-layer and multilayer neural networks. As a result of comparative analysis of algorithms for extracting primary biometric features and comparing the generated image to a private key within the proposed authentication system, it was found that deep convolutional networks and neural network bidirectional associative memory are the most effective approach to process the data. In the research, an approach based on the integration of a biometric system and a cryptographic module was proposed, which allows using of a generated secret cryptographic key based on a biometric sample as the output of a neural network. The RSA algorithm is chosen to generate a private cryptographic key by use of convolutional neural networks and Python libraries. The software authentication module is implemented based on the client-server architecture using various internal Python libraries. Such authentication system should be used in systems where the user data and his valuable information resources are stored or where the user can perform certain valuable operations for which a cryptographic key is required. Proposed software module based on convolutional neural networks will be a perfect tool for ensuring the confidentiality of information and for all information-communication systems, because protecting information system from unauthorized access is one of the most pressing problems. This approach as software module solves the problem of secure generating and storing the secret key and author propose combination of the convolutional neural network with bidirectional associative memory, which is used to recognize the biometric sample, generate the image, and match it with a cryptographic key. The use of this software approach allows today to reduce the probability of errors of the first and second kind in authentication system and absolute number of errors was minimized by an average of 1,5 times. The proportion of correctly recognized images by the comparating together convolutional networks and neural network bidirectional associative memory in the authentication software module increased to 96,97%, which is on average from 1,08 times up to 1,01 times The authors further plan a number of scientific and technical solutions to develop and implement effective methods, tools to meet the requirements, principles and approaches to cybersecurity and cryptosystems for provide integrity and onfidentiality of information in experimental computer systems and networks.

Segmentation of semantic and perceptual boundaries affects associative memory and risk of false memories

Temporally structured sequences of experiences, such as narratives or life events, are segmented in memory into discrete situational models. In segmentation, contextual shifts are processed as situational boundaries that temporally cluster items according to the perceived contexts. As such, segmentation enhances associative binding of items within a situational model. One side effect of enhanced associative processing is increased risk of false recollections for not-presented, semantically related items. If so, do boundaries facilitate false recollections, or does segmentation protect against them? In two experiments, we introduced situational shifts in word sequences in the form of semantic and perceptual boundaries, with semantic relatedness between words or the frame color around a word changing on a regular basis. After encoding, we tested participants’ associative memory performance and false recollection rates. In Experiment 1, color boundaries occurred synchronously or asynchronously to semantic boundaries. We found better associative recognition, but also more false recollections, for synchronous than asynchronous boundaries. In Experiment 2, color boundaries occurred synchronous to semantic boundaries or were absent entirely. We found that false recollection rates elicited by semantic boundaries increased when color boundaries were absent. We also tested associative memory performance using a non-semantic, temporal memory task. We found better temporal memory performance for semantic boundaries, as well as a negative correlation between increased false recollection rates and better temporal memory performance for semantic lists, but not for random lists. We discuss implications for false memory theories and segmentation of narrative materials in false memory research.

An Empirical Analysis of Corporate Financial Management Risk Prediction Based on Associative Memory Neural Network

As a human brain-like computational model that can reflect the cognitive function of the brain, the problem of dynamic analysis of associative memory neural networks has attracted the attention of scholars. This paper combines associative memory neural networks with enterprise financial management risks, studies the synchronization control and stability analysis problems of unidirectional associative memory-like human brain amnestic neural networks with perturbation and mixed time-varying time lags, proposes a bidirectional associative memory-like brain stochastic amnestic neural network model with mixed time-varying time lags, designs a discrete-time sampling control strategy based on the model, and studies various types of recent financial risks. Based on the early warning research, based on the associative memory neural network method, we propose to reconstruct the risk categories, including improving the enterprise risk management system, enhancing the awareness of financial risk management from top to bottom, and strengthening the core competitiveness of the enterprise itself and control measures for financing, investment, operation, and cash flow risks.