Search

In this study, we propose the use of artificial synaptic transistors with coplanar-gate structures fabricated on paper substrates comprising biocompatible and low-cost potato-starch electrolyte and indium–gallium–zinc oxide (IGZO) channels. The electrical double layer (EDL) gating effect of potato-starch electrolytes enabled the emulation of biological synaptic plasticity. Frequency dependence measurements of capacitance using a metal-insulator-metal capacitor configuration showed a 1.27 μF/cm2 at a frequency of 10 Hz. Therefore, strong capacitive coupling was confirmed within the potato-starch electrolyte/IGZO channel interface owing to EDL formation because of internal proton migration. An electrical characteristics evaluation of the potato-starch EDL transistors through transfer and output curve resulted in counterclockwise hysteresis caused by proton migration in the electrolyte; the hysteresis window linearly increased with maximum gate voltage. A synaptic functionality evaluation with single-spike excitatory post-synaptic current (EPSC), paired-pulse facilitation (PPF), and multi-spike EPSC resulted in mimicking short-term synaptic plasticity and signal transmission in the biological neural network. Further, channel conductance modulation by repetitive presynaptic stimuli, comprising potentiation and depression pulses, enabled stable modulation of synaptic weights, thereby validating the long-term plasticity. Finally, recognition simulations on the Modified National Institute of Standards and Technology (MNIST) handwritten digit database yielded a 92% recognition rate, thereby demonstrating the applicability of the proposed synaptic device to the neuromorphic system. [ABSTRACT FROM AUTHOR]

Biomimetic neuromorphic sensing systems, inspired by the structure and function of biological neural networks, represent a major advancement in the field of sensing technology and artificial intelligence. This review paper focuses on the development and application of electrolyte gated transistors (EGTs) as the core components (synapses and neuros) of these neuromorphic systems. EGTs offer unique advantages, including low operating voltage, high transconductance, and biocompatibility, making them ideal for integrating with sensors, interfacing with biological tissues, and mimicking neural processes. Major advances in the use of EGTs for neuromorphic sensory applications such as tactile sensors, visual neuromorphic systems, chemical neuromorphic systems, and multimode neuromorphic systems are carefully discussed. Furthermore, the challenges and future directions of the field are explored, highlighting the potential of EGT-based biomimetic systems to revolutionize neuromorphic prosthetics, robotics, and human–machine interfaces. Through a comprehensive analysis of the latest research, this review is intended to provide a detailed understanding of the current status and future prospects of biomimetic neuromorphic sensory systems via EGT sensing and integrated technologies. [ABSTRACT FROM AUTHOR]

This paper presents a neural network predictive control scheme for studying the coagulation process of wastewater treatment in a paper mill. A multi-layer back-propagation neural network is employed to model the nonlinear relationships between the removal rates of pollutants and the chemical dosages, in order to adapt the system to a variety of operating conditions and acquire a more flexible learning ability. The system includes a neural network emulator of the reaction process, a neural network controller, and an optimization procedure based on a performance function that is used to identify desired control inputs. The gradient descent algorithm method is used to realize the optimization procedure. The results indicate that reasonable forecasting and control performances have been achieved through the developed system. [Copyright &y& Elsevier]

Recurrent neural networks (RNNs) transmit information over time through recurrent connections. In contrast, biological neural networks use many other temporal processing mechanisms. One of these mechanisms is the inter-neuron delays caused by varying axon properties. Recently, this feature was implemented in echo state networks (ESNs), a type of RNN, by assigning spatial locations to neurons and introducing distance-dependent inter-neuron delays. These delays were shown to significantly improve ESN task performance. However, thus far, it is still unclear why distance-based delay networks (DDNs) perform better than ESNs. In this paper, we show that by optimizing inter-node delays, the memory capacity of the network matches the memory requirements of the task. As such, networks concentrate their memory capabilities to the points in the past which contain the most information for the task at hand. Moreover, we show that DDNs have a greater total linear memory capacity, with the same amount of non-linear processing power. [ABSTRACT FROM AUTHOR]

Owing to the highly parallel network structure of the biological neural network and its triggered processing mode, tactile sensory neurons can realize the perception of external signals and the functions of perception, memory, and data processing by adjusting the synaptic weight. In this paper, a piezoresistive pressure sensor is combined with a memristor to design an artificial tactile sensory neuron. The polyurethane sponge sensor has excellent sensitivity and can convert physical stimuli into electrical signals, and the chitosan‐based memristor has stable bipolar resistive switching characteristics, allowing further information to be memorized and processed. The neuron can respond to tactile stimuli of different degrees, durations, and frequencies; realize potentiation/depression modulation, paired‐pulse facilitation, and spike‐timing‐dependent plasticity; exhibit spike‐rate‐dependent plasticity; and store and erase tactile information through memistor state switching, which has great application potential in biological sensing systems. [ABSTRACT FROM AUTHOR]

The functional networks of the human brain exhibit the structural characteristics of a scale-free topology, and these neural networks are exposed to the electromagnetic environment. In this paper, we consider the effects of magnetic induction on synchronous activity in biological neural networks, and the magnetic effect is evaluated by the four-stable discrete memristor. Based on Rulkov neurons, a scale-free neural network model is established. Using the initial value and the strength of magnetic induction as control variables, numerical simulations are carried out. The research reveals that the scale-free neural network exhibits multiple coexisting behaviors, including resting state, period-1 bursting synchronization, asynchrony, and chimera states, which are dependent on the different initial values of the multi-stable discrete memristor. In addition, we observe that the strength of magnetic induction can either enhance or weaken the synchronization in the scale-free neural network when the parameters of Rulkov neurons in the network vary. This investigation is of significant importance in understanding the adaptability of organisms to their environment. [ABSTRACT FROM AUTHOR]

Traditional biological neural networks cannot simulate the real situation of the abrupt synaptic connections between neurons while modeling associative memory of human brains. In this paper, the memristive multidirectional associative memory neural networks (MAMNNs) with mixed time-varying delays are investigated in the sense of Filippov solution. First, three steps are given to prove the existence of the almost periodic solution. Two new lemmas are proposed to prove the boundness of the solution and the asymptotical almost periodicity of the solution by constructing Lyapunov function. Second, the uniqueness and global exponential stability of the almost periodic solution of memristive MAMNNs are investigated by a new Lyapunov function. The sufficient conditions guaranteeing the properties of almost periodic solution are derived based on the relevant definitions, Halanay inequality and Lyapunov function. The investigation is an extension of the research on the periodic solution and almost periodic solution of bidirectional associative memory neural networks. Finally, numerical examples with simulations are presented to show the validity of the main results. [ABSTRACT FROM AUTHOR]

Multifunctional biological neural networks exploit multistability in order to perform multiple tasks without changing any network properties. Enabling artificial neural networks (ANNs) to obtain certain multistabilities in order to perform several tasks, where each task is related to a particular attractor in the network's state space, naturally has many benefits from a machine learning perspective. Given the association to multistability, in this paper, we explore how the relationship between different attractors influences the ability of a reservoir computer (RC), which is a dynamical system in the form of an ANN, to achieve multifunctionality. We construct the "seeing double" problem in order to systematically study how a RC reconstructs a coexistence of attractors when there is an overlap between them. As the amount of overlap increases, we discover that for multifunctionality to occur, there is a critical dependence on a suitable choice of the spectral radius for the RC's internal network connections. A bifurcation analysis reveals how multifunctionality emerges and is destroyed as the RC enters a chaotic regime that can lead to chaotic itinerancy. [ABSTRACT FROM AUTHOR]

In order to enrich the dynamic behaviors of discrete neuron models and more effectively mimic biological neural networks, this paper proposes a bistable locally active discrete memristor (LADM) model to mimic synapses. We explored the dynamic behaviors of neural networks by introducing the LADM into two identical Rulkov neurons. Based on numerical simulation, the neural network manifested multistability and new firing behaviors under different system parameters and initial values. In addition, the phase synchronization between the neurons was explored. Additionally, it is worth mentioning that the Rulkov neurons showed synchronization transition behavior; that is, anti-phase synchronization changed to in-phase synchronization with the change in the coupling strength. In particular, the anti-phase synchronization of different firing patterns in the neural network was investigated. This can characterize the different firing behaviors of coupled homogeneous neurons in the different functional areas of the brain, which is helpful to understand the formation of functional areas. This paper has a potential research value and lays the foundation for biological neuron experiments and neuron-based engineering applications. [ABSTRACT FROM AUTHOR]

Leaky Integrator Echo State Network (Leaky-ESN) is a useful training method for handling time series prediction problems. However, the singular coupling of all neurons in the reservoir makes Leaky-ESN less effective for sophisticated learning tasks. In this paper, we propose a new improvement to the Leaky-ESN model called the Multiple-Reservoir Hierarchical Echo State Network (MH-ESN). By introducing a new mechanism for constructing the reservoir, the efficiency of the network in handling training tasks is improved. The hierarchical structure is used in the process of constructing the reservoir mechanism of MH-ESN. The MH-ESN consists of multiple layers, each comprising a multi-reservoir echo state network model. The sub-reservoirs within each layer are linked via principal neurons, which mimics the functioning of a biological neural network. As a result, the coupling among neurons in the reservoir is decreased, and the internal dynamics of the reservoir are improved. Based on the analysis results, the MH-ESN exhibits significantly better prediction accuracy than Leaky-ESN for complex time series prediction. [ABSTRACT FROM AUTHOR]

In this paper we present computational simulations of a mathematical model describing the time evolution of membrane potentials in a GABAergic neural network. This model, with stochastic and evolutionary characteristics, is an application of the version introduced previously where the authors present the continuous time version of a new class of stochastic models for biological neural networks. The goal is to computationally simulate the model (with the interaction conditions of a GABAergic network) and make biological inferences. More specifically, the computational simulations of the model that describe spiking neurons with electrophysiological characteristics of a brain region called substantia nigra pars reticulata, emphasize changes in desynchronized firing activity and how changes in individual activity propagate through the network. [ABSTRACT FROM AUTHOR]

Influenza affects millions of people every year. It causes a considerable amount of medical visits and hospitalisations as well as hundreds of thousands of deaths. Forecasting influenza prevalence with good accuracy can significantly help public health agencies to timely react to seasonal or novel strain epidemics. Although significant progress has been made, influenza forecasting remains a challenging modelling task. In this paper, we propose a methodological framework that improves over the state-of-the-art forecasting accuracy of influenza-like illness (ILI) rates in the United States. We achieve this by using Web search activity time series in conjunction with historical ILI rates as observations for training neural network (NN) architectures. The proposed models incorporate Bayesian layers to produce associated uncertainty intervals to their forecast estimates, positioning themselves as legitimate complementary solutions to more conventional approaches. The best performing NN, referred to as the iterative recurrent neural network (IRNN) architecture, reduces mean absolute error by 10.3% and improves skill by 17.1% on average in nowcasting and forecasting tasks across 4 consecutive flu seasons. Author summary: Modelling the prevalence of an infectious disease enables public health organisations to prepare for and minimise its impact. However, traditional disease indicators are often quite restrictive as they provide information with a significant delay. Recent research efforts have provided evidence of the value of alternative information sources such as Web search activity trends. Our work incorporates this information into influenza forecasting models to achieve state-of-the-art accuracy. In addition, the proposed Bayesian neural network architectures also provide associated uncertainty estimates for the forecasts, positioning our methodology as a practical complementary tool for disease surveillance and policy making. [ABSTRACT FROM AUTHOR]

Currently, livestock and poultry farming is gradually developing towards modernization and scale, and closed livestock and poultry farms are widely used for poultry feeding management, but at the same time, the farming risks of large-scale farms are increasing. In this paper, based on the study of wireless sensor networks and biological neural network models, the environmental factors that mainly affect the growth of domestic rabbits are analyzed as an example, and the technology is used to design and implement an environmental monitoring system for modern farms. The design of the system is divided into three main parts: hardware design of each node, software design, and upper computer monitoring software design. The hardware part of the system uses coordinator nodes, router nodes, sensor nodes, and control nodes to form a wireless sensor network in the farm, carries out the hardware circuit design of each node, and based on the protocol stack, designs the software program of each node to realize the collection, transmission, and regulation of environmental information in the farm. In the upper computer part, the design and development of the upper computer monitoring software interface are used to complete the real-time display of environmental data, historical query, database storage, and curve drawing, and to design a remote client data query system based on the architecture to realize the query of environmental data of the farm by remote users and to carry out monitoring fault intelligent identification alarm. At the same time, the paper investigates the optimal deployment of wireless sensor network nodes and searches for the optimal location of sensor nodes through an improved biological neural network algorithm to maximize the network coverage and reduce the coverage of blind areas, and conducts simulation experiments with the coverage rate of a rabbit farm as the optimization target. [ABSTRACT FROM AUTHOR]

In this paper, a comprehensive quantitative and biological neural network optimization model of sports industry structure is thoroughly studied and analyzed using knowledge graphs. To address the problems of poor performance interpretability deficiency of knowledge graph-based recommendation methods in the face of relational sparse graphs, a pretraining-based implicit characterization algorithm strategy is proposed for the recall stage, which can solve the problems of difficulty in going online and high delay in the recall stage of the recommendation system while improving the accuracy, and not only this can be applied in the recall stage, but also the sorting and postsorting modules can be used as features. To study the relationship between signaling activity and energy metabolism of pyramidal neurons, an empirical model of the synaptic vesicle cycle is proposed to simulate the synaptic transmission process, the role played by energy metabolism in synaptic transmission is studied from the perspective of feedback control, and the quantitative relationship between neuronal pulse discharge frequency, energy consumption, and information quantity in dendritic integration is analyzed using the cable theory and atrial chamber model. It was found that, when 0 ≤ ε ≤ 0.6 , the chaotic region shrinks and eventually disappears with the increase of the memory factor ε ; however, when 0.6 ≤ ε ≤ 1 is used, chaos is recreated and the chaotic area gradually increases with the increase of the memory factor ε. This paper conducts comparative experiments on data sets in the recommendation domain and verifies that the proposed model and the feature intersection module can effectively perform feature interaction between items and entities, thus enhancing the recommendation effect. [ABSTRACT FROM AUTHOR]

• Physical behaviors of memristive systems can be related to the bio-physical dynamics of biological neural elements. • Spiking behaviors subject to input stimuli of LIF neurons made of memristive elements can be extrapolated to develop on-chip learning algorithms. • Rate coding is a viable alternative to temporal or population coding for in-hardware SNNs. • It is possible to perform a non-linear functions like XOR using a single neuron in SNNs. • A hybrid approach relying on ANN like gradient calculation can be used to learn in SNNs. Neuromorphic computing is a novel computing paradigm that aims to mimic the behavior of biological neural networks for efficiently solving complex problems. While CMOS based neurons and synapses have been developed, they are limited in their ability to demonstrate bio-realistic dynamics. This, coupled with the fact that a huge number of these individual devices are required to build neurons and synapses, limits the scaling and power efficiency of such systems. A viable answer to this problem is neuromemristive systems that are based on memristor devices. These devices exhibit physical behaviors that can be related to the bio-physical dynamics of synapses and neurons. In this paper, a rate-coded all memristive "spiking neural network" (SNN) is presented. The proposed SNN is built with an active memristor neuron based on vanadium dioxide (VO 2) coupled with a non-volatile memristor synapse. The results are validated by first simulating spiking versions of two Boolean functions viz., AND and XOR gates in SPICE. With features extracted from the small neural nets, a large-scale 3-layer spiking neural network is then simulated in Python which yields a validation accuracy of 87% on the MNIST dataset of handwritten digits. One of the prime features of this work is the realization of the XOR function using a single neuron which is not possible without the use of 2-layers of neurons in traditional neural networks. Another significant contribution is the utilization of a gradient-based learning approach for online training of a large-scale SNN. For this, we use the inherent activation function (Sigmoid/ReLU) of the proposed neuron design. [ABSTRACT FROM AUTHOR]

Tongue and its movements can be used for several medical-related tasks, such as identifying a disease and tracking a rehabilitation. To be able to focus on a tongue region, the tongue segmentation is needed to compute a region of interest for a further analysis. This paper proposes an encoder-decoder CNN-based architecture for segmenting a tongue in an image. The encoder module is mainly used for the tongue feature extraction, while the decoder module is used to reconstruct a segmented tongue from the extracted features based on training images. In addition, the residual multi-kernel pooling (RMP) is also applied into the proposed network to help in encoding multiple scales of the features. The proposed method is evaluated on two publicly available datasets under a scenario of front view and one tongue posture. It is then tested on a newly collected dataset of five tongue postures. The reported performances show that the proposed method outperforms existing methods in the literature. In addition, the re-training process could improve applying the trained model on unseen dataset, which would be a necessary step of applying the trained model on the real-world scenario. [ABSTRACT FROM AUTHOR]

With the rapid development of economy and the increasing improvement of agricultural production level, people's demand for fruits is also increasing year by year. China is the largest fruit production and consumption country in the world. According to relevant statistics reported for China, by the end of 2019, the total amount of various fruits sold had reached about 270 million tons, with apples accounting for 48% of the global output and pears accounting for 69% of the national total output. However, China's fruit picking is still dominated by manual picking process, which takes a lot of manpower and time to complete, therefore resulting in low fruit picking efficiency. Some farmers are unable to complete fruit picking in a short time, resulting in a large number of fruits unable to be listed, resulting in huge losses. To solve this problem, this paper focuses on the visual recognition and path planning for intelligent fruit-picking robot. Using robot to complete fruit picking is the best way at present. This paper establishes a picking robot recognition and positioning system based on stereo vision, which is used to identify and locate the fruits planted in the orchard area. The coordinate error of the target point of the intelligent fruit-picking robot coordinate system is less than 10 mm, which has high accuracy. Then, the path of intelligent fruit-picking robot is planned based on visual feedback algorithm and biological stimulation neural network. Our empirical evaluations suggest that the proposed robot walks in the planting park in the shape of "zigzag" and realizes full-coverage path planning after 6 turns. The results show the efficiency of the intelligent method. [ABSTRACT FROM AUTHOR]

The performance of the recently proposed orientational beamforming (OBF) system for ultra-wideband (UWB) signals degrades seriously in multiuser situation. Although various techniques have been proposed in the literature to suppress multiuser interference in the time-hopping UWB system, there has been no research on interference suppression in the OBF UWB system. In this paper, we propose a nonlinear OBF system via a modified radial basis function (RBF) neural network, named the RBF-OBF system, to reject orientational UWB interferences caused by multiuser transmission in the OBF system. The RBF neural network is selected mainly because of its simple structure, convenient training process, and fast convergence speed. However, the conventional Euclidean-distance-based RBF neural network is not suitable for the training process used in this work, as the training samples are noiseless orientational steering vectors in frequency domain. Therefore, a modified RBF neural network is proposed in this paper, which evaluates the similarity between an input vector and the center vector of a hidden layer neuron by their signed complex correlation coefficient. Numerous simulations demonstrate that compared to the conventional OBF system, the proposed RBF-OBF system can significantly reduce the bit error rates (BERs) under different noise and interference situations. The BER reduction rate of the proposed RBF-OBF system is higher than 92% in additive white Gaussian noise channel and higher than 89% in line-of-sight multipath channel. [ABSTRACT FROM AUTHOR]

Electroencephalogram (EEG) data are produced in quantity for measuring brain activity in response to external stimuli. With the rapid development of brain-inspired intelligence, spiking neural network (SNN) possesses the potential to handle EEG data by using spiking activity transmitted among spatially located synapses and neurons. As an original and unifying SNN architecture, NeuCube, is developed to model, recognize and understand EEG data. However, the NeuCube still faces some challenges for EEG-based pattern recognition, such as few labeled data and changes of data probability distribution. Hence, this paper proposes a novel method to improve the performance of the NeuCube for EEG-based pattern recognition by transfer learning. In the first place, the covariance matrix alignment of EEG data is implemented for every subject in the Euclidean space, which reduces the probability distribution discrepancy of EEG data between different subjects. Different estimation methods for reference covariance matrix are tested and the optimal one is selected for different subjects. Secondly, spatio-temporal features of EEG data are extracted based on the NeuCube reservoir. Since hyper-parameters of the NeuCube reservoir have a great impact on its spatio-temporal representation, an improved cuckoo search algorithm is proposed to discover the optimal hyper-parameters for obtaining the optimal spatio-temporal features. Last but not least, a weighted transfer support vector machine is proposed to improve the original output classifier of the NeuCube in order to make the model adaptive to the cross-domain variability of EEG data. The proposed method is tested on open dataset 2a from BCI competition IV 2008 and achieves good spatio-temporal pattern recognition results. Furthermore, the neuron connectivity and activation level associated with the process of mental tasks are illustrated. [ABSTRACT FROM AUTHOR]

The main advantages of spiking neural networks are the high biological plausibility and their fast response due to spiking behaviour. The response time decreases significantly in the hardware implementation of SNN because the neurons operate in parallel. Compared with the traditional computational neural network, the SNN use a lower number of neurons, which also reduces their cost. Another critical characteristic of SNN is their ability to learn by event association that is determined mainly by postsynaptic mechanisms such as long-term potentiation. However, in some conditions, presynaptic plasticity determined by post-tetanic potentiation occurs due to the fast activation of presynaptic neurons. This violates the Hebbian learning rules that are specific to postsynaptic plasticity. Hebbian learning improves the SNN ability to discriminate the neural paths trained by the temporal association of events, which is the key element of learning in the brain. This paper quantifies the efficiency of Hebbian learning as the ratio between the LTP and PTP effects on the synaptic weights. On the basis of this new idea, this work evaluates for the first time the influence of the number of neurons on the PTP/LTP ratio and consequently on the Hebbian learning efficiency. The evaluation was performed by simulating a neuron model that was successfully tested in control applications. The results show that the firing rate of postsynaptic neurons p o s t depends on the number of presynaptic neurons p r e , which increases the effect of LTP on the synaptic potentiation. When p o s t activates at a requested rate, the learning efficiency varies in the opposite direction with the number of p r e s , reaching its maximum when fewer than two p r e s are used. In addition, Hebbian learning is more efficient at lower presynaptic firing rates that are divisors of the target frequency of p o s t . This study concluded that, when the electronic neurons additionally model presynaptic plasticity to LTP, the efficiency of Hebbian learning is higher when fewer neurons are used. This result strengthens the observations of our previous research where the SNN with a reduced number of neurons could successfully learn to control the motion of robotic fingers. [ABSTRACT FROM AUTHOR]

A biological neural network in the cortex forms a neural field. Neurons in the field have their own receptive fields, and connection weights between two neurons are random but highly correlated when they are in close proximity in receptive fields. In this paper, we investigate such neural fields in a multilayer architecture to investigate the supervised learning of the fields. We empirically compare the performances of our field model with those of randomly connected deep networks. The behavior of a randomly connected network is investigated on the basis of the key idea of the neural tangent kernel regime, a recent development in the machine learning theory of over-parameterized networks; for most randomly connected neural networks, it is shown that global minima always exist in their small neighborhoods. We numerically show that this claim also holds for our neural fields. In more detail, our model has two structures: (i) each neuron in a field has a continuously distributed receptive field, and (ii) the initial connection weights are random but not independent, having correlations when the positions of neurons are close in each layer. We show that such a multilayer neural field is more robust than conventional models when input patterns are deformed by noise disturbances. Moreover, its generalization ability can be slightly superior to that of conventional models. [ABSTRACT FROM AUTHOR]

Background: Building biological networks with a certain function is a challenge in systems biology. For the functionality of small (less than ten nodes) biological networks, most methods are implemented by exhausting all possible network topological spaces. This exhaustive approach is difficult to scale to large-scale biological networks. And regulatory relationships are complex and often nonlinear or non-monotonic, which makes inference using linear models challenging. Results: In this paper, we propose a multi-layer perceptron-based differential equation method, which operates by training a fully connected neural network (NN) to simulate the transcription rate of genes in traditional differential equations. We verify whether the regulatory network constructed by the NN method can continue to achieve the expected biological function by verifying the degree of overlap between the regulatory network discovered by NN and the regulatory network constructed by the Hill function. And we validate our approach by adapting to noise signals, regulator knockout, and constructing large-scale gene regulatory networks using link-knockout techniques. We apply a real dataset (the mesoderm inducer Xenopus Brachyury expression) to construct the core topology of the gene regulatory network and find that Xbra is only strongly expressed at moderate levels of activin signaling. Conclusion: We have demonstrated from the results that this method has the ability to identify the underlying network topology and functional mechanisms, and can also be applied to larger and more complex gene network topologies. [ABSTRACT FROM AUTHOR]

Spiking neural networks (SNNs) have shown clear advantages over traditional artificial neural networks (ANNs) for low latency and high computational efficiency, due to their event-driven nature and sparse communication. However, the training of deep SNNs is not straightforward. In this paper, we propose a novel ANN-to-SNN conversion and layer-wise learning framework for rapid and efficient pattern recognition, which is referred to as progressive tandem learning. By studying the equivalence between ANNs and SNNs in the discrete representation space, a primitive network conversion method is introduced that takes full advantage of spike count to approximate the activation value of ANN neurons. To compensate for the approximation errors arising from the primitive network conversion, we further introduce a layer-wise learning method with an adaptive training scheduler to fine-tune the network weights. The progressive tandem learning framework also allows hardware constraints, such as limited weight precision and fan-in connections, to be progressively imposed during training. The SNNs thus trained have demonstrated remarkable classification and regression capabilities on large-scale object recognition, image reconstruction, and speech separation tasks, while requiring at least an order of magnitude reduced inference time and synaptic operations than other state-of-the-art SNN implementations. It, therefore, opens up a myriad of opportunities for pervasive mobile and embedded devices with a limited power budget. [ABSTRACT FROM AUTHOR]

In this paper, we propose a novel class of networks named as quaternion-valued inertial memristor-based neural networks (QVIMNNs) by introducing inertial term and memristor into traditional quaternion-valued neural networks (QVNNs). The problem of fixed-time synchronization of the QVIMNNs is investigated based on the variable transformation and Lyapunov functional method. It is shown that two types of activation functions are considered and a novel criteria guaranteeing fixed-time synchronization for each cases are then achieved by designing different types of controllers. This paper attempts to pave a new way to investigate neural networks classes with numerical simulations support to demonstrate the correctness of the obtained results. [ABSTRACT FROM AUTHOR]

Big data is a massive and diverse form of unstructured data, which needs proper analysis and management. It is another great technological revolution after the Internet, the Internet of Things, and cloud computing. This paper firstly studies the related concepts and basic theories as the origin of research. Secondly, it analyzes in depth the problems and challenges faced by Chinese government management under the impact of big data. Again, we explore the opportunities that big data brings to government management in terms of management efficiency, administrative capacity, and public services and believe that governments should seize opportunities to make changes. Brainlike computing attempts to simulate the structure and information processing process of biological neural network. This paper firstly analyzes the development status of e-government at home and abroad, studies the service-oriented architecture (SOA) and web services technology, deeply studies the e-government and SOA theory, and discusses this based on the development status of e-government in a certain region. Then, the deep learning algorithm is used to construct the monitoring platform to monitor the government behavior in real time, and the deep learning algorithm is used to conduct in-depth mining to analyze the government's intention behavior. [ABSTRACT FROM AUTHOR]

In this paper, BP neural network is used to collect header height, AMEsim is used to simulate and analyze header height adjustment hydraulic system, and fuzzy PID control is used to adjust header lifting hydraulic cylinder to stabilize header height. The experimental results of harvesting different crops show that under the header height automatic control system, the error between the actual height of crop harvesting and the set height is within 15 mm, and the harvesting effect is good, which can meet the automatic regulation requirements of the header height of the multi crop combine harvester. [ABSTRACT FROM AUTHOR]

Most deep learning models for temporal regression directly output the estimation based on single input images, ignoring the relationships between different images. In this paper, we propose deep relation learning for regression, aiming to learn different relations between a pair of input images. Four non-linear relations are considered: “cumulative relation,” “relative relation,” “maximal relation” and “minimal relation.” These four relations are learned simultaneously from one deep neural network which has two parts: feature extraction and relation regression. We use an efficient convolutional neural network to extract deep features from the pair of input images and apply a Transformer for relation learning. The proposed method is evaluated on a merged dataset with 6,049 subjects with ages of 0–97 years using 5-fold cross-validation for the task of brain age estimation. The experimental results have shown that the proposed method achieved a mean absolute error (MAE) of 2.38 years, which is lower than the MAEs of 8 other state-of-the-art algorithms with statistical significance (p<0.05) in paired T-test (two-side). [ABSTRACT FROM AUTHOR]

Simple Summary: The brain is an incredibly complex system possessing outstanding abilities to perform difficult tasks through a vast number of densely interconnected neurons. Aimed at discovering the underlying mechanisms of the brain, a number of spiking neural networks have been proposed to mimic biological neural dynamics. Subsequently, to perceive how the neural networks in the brain work, simulation and hardware realization of large-scale systems, similar to the brain, is an essential requirement. Behavior of a single neuron can be described by the mathematical equations in different levels of computing and biological accuracy. In this approach, a new modified ADEX model is presented based on sampling frequency by the nonlinear functions of the original model. This new model is capable for reproducing all aspects of the original model in low-error and high-degree of similarity conditions. Finally, the proposed model can be implemented on digital hardware platforms to have a real digital system. Digital results show the increase in system speed (frequency) and overall saving in hardware resources (compared by the original model and other similar works). This low-cost digital hardware is applied in large-scale neuronal networks. Design and implementation of biological neural networks is a vital research field in the neuromorphic engineering. This paper presents LUT-based modeling of the Adaptive Exponential integrate-and-fire (ADEX) model using Nyquist frequency method. In this approach, a continuous term is converted to a discrete term by sampling factor. This new modeling is called N-LUT-ADEX (Nyquist-Look Up Table-ADEX) and is based on accurate sampling of the original ADEX model. Since in this modeling, the high-accuracy matching is achieved, it can exactly reproduce the spiking patterns, which have the same behaviors of the original neuron model. To confirm the N-LUT-ADEX neuron, the proposed model is realized on Virtex-II Field-Programmable Gate Array (FPGA) board for validating the final hardware. Hardware implementation results show the high degree of similarity between the proposed and original models. Furthermore, low-cost and high-speed attributes of our proposed neuron model will be validated. Indeed, the proposed model is capable of reproducing the spiking patterns in terms of low overhead costs and higher frequencies in comparison with the original one. The properties of the proposed model cause can make it a suitable choice for neuromorphic network implementations with reduced-cost attributes. [ABSTRACT FROM AUTHOR]

In this paper, robust exponential synchronization of memristor-based neural networks (MNNs) with nonidentical characteristics is investigated. Coefficient mismatch, time-varying delay mismatch, and activation function mismatch are considered between the drive and the response MNNs. Pinning control strategy is developed to realize robust exponential synchronization and the stability criteria is established by using the Lyapunov function method and differential inclusion theory. Furthermore, the stable region of controller parameters is computed to guarantee that the synchronization errors enter a predetermined error bound within given settling time. Finally, the effectiveness of the proposed methods is verified by the numerical simulations. The methods presented in this paper offer novel schemes for robust exponential synchronization of nonidentical MNNs. [ABSTRACT FROM AUTHOR]

Self-repair is fundamental to biological neural networks. In a neural network with a large number of elements, the probability of failure of each component increases. With the breakdown of each part, there may be a significant difference in the final results that will be completely affected by this defect. The existence of a process for detecting the error and compensating it by recruiting healthy elements leads to improved performance. This is where adjacent synapses proxy faulty synapses to avoid disturbances in the network function, thereby compensating the incurred error. In the present research, a self-repairing analog circuit is designed based on an astrocyte-neuron interaction and new synapse architecture. The designed circuit builds upon a software model of an astrocyte-neuron network with the proven ability to detect errors and undertake self-repair. The results obtained from our circuit show that, when an error occurs in the synapses associated with a neuron, the currents within functioning synapses of the same neuron increase. This increase is made by receiving feedback from adjacent astrocytes and other synapses. The process maintains the network function, compensating incurred errors in the network, presenting a neural network-based analog circuit with self-repairing capability, while considering the effect of astrocytes. In this paper, extensive simulation results using HSPICE with 0.35 μm CMOS technology are provided for the evaluation of the proposed circuit. [ABSTRACT FROM AUTHOR]

Live fluorescence imaging has demonstrated the dynamic nature of dendritic spines, with changes in shape occurring both during development and in response to activity. The structure of a dendritic spine correlates with its functional efficacy. Learning and memory studies have shown that a great deal of the information stored by a neuron is contained in the synapses. High precision tracking of synaptic structures can give hints about the dynamic nature of memory and help us understand how memories evolve both in biological and artificial neural networks. Experiments that aim to investigate the dynamics behind the structural changes of dendritic spines require the collection and analysis of large time-series datasets. In this paper, we present an open-source software called SpineS for automatic longitudinal structural analysis of dendritic spines with additional features for manual intervention to ensure optimal analysis. We have tested the algorithm on in-vitro, in-vivo, and simulated datasets to demonstrate its performance in a wide range of possible experimental scenarios. [ABSTRACT FROM AUTHOR]

The past decade has witnessed the great success of deep neural networks in various domains. However, deep neural networks are very resource-intensive in terms of energy consumption, data requirements, and high computational costs. With the recent increasing need for the autonomy of machines in the real world, e.g., self-driving vehicles, drones, and collaborative robots, exploitation of deep neural networks in those applications has been actively investigated. In those applications, energy and computational efficiencies are especially important because of the need for real-time responses and the limited energy supply. A promising solution to these previously infeasible applications has recently been given by biologically plausible spiking neural networks. Spiking neural networks aim to bridge the gap between neuroscience and machine learning, using biologically realistic models of neurons to carry out the computation. Due to their functional similarity to the biological neural network, spiking neural networks can embrace the sparsity found in biology and are highly compatible with temporal code. Our contributions in this work are: (i) we give a comprehensive review of theories of biological neurons; (ii) we present various existing spike-based neuron models, which have been studied in neuroscience; (iii) we detail synapse models; (iv) we provide a review of artificial neural networks; (v) we provide detailed guidance on how to train spike-based neuron models; (vi) we revise available spike-based neuron frameworks that have been developed to support implementing spiking neural networks; (vii) finally, we cover existing spiking neural network applications in computer vision and robotics domains. The paper concludes with discussions of future perspectives. [ABSTRACT FROM AUTHOR]

Neural Network Coding and Representation (NNR) is the first international standard for efficient compression of neural networks (NNs). The standard is designed as a toolbox of compression methods, which can be used to create coding pipelines. It can be either used as an independent coding framework (with its own bitstream format) or together with external neural network formats and frameworks. For providing the highest degree of flexibility, the network compression methods operate per parameter tensor in order to always ensure proper decoding, even if no structure information is provided. The NNR standard contains compression-efficient quantization and deep context-adaptive binary arithmetic coding (DeepCABAC) as core encoding and decoding technologies, as well as neural network parameter pre-processing methods like sparsification, pruning, low-rank decomposition, unification, local scaling and batch norm folding. NNR achieves a compression efficiency of more than 97% for transparent coding cases, i.e. without degrading classification quality, such as top-1 or top-5 accuracies. This paper provides an overview of the technical features and characteristics of NNR. [ABSTRACT FROM AUTHOR]

This paper presents a method for multi-label classification on source EEG signal datasets using DCNN and GPReLU with various signal patches. Performance has been appraised using ROC Curves for multi-label classification. Emotion is a phenomenal neurological expression that releases bio-signals with electrical voltage in the brain, which is read through as EEG signals. Reading the EEG signals and analyzing them is carried out for many purposes in science and technology. A common DCNN can simply classify the signal datasets, which often is inaccurate, although classified signal entities may contain features of other classes. Classification in uncertain data leads to class imbalances. Therefore, multi-label classification suffices the need for data analysis which anatomizes the EEG signal data as containing features of many categories. A DCNN with Geometric Parameterized ReLU is introduced to smoothen the pooling activity and work swiftly on various image patches of the source EEG signal image datasets. [ABSTRACT FROM AUTHOR]

Bio‐inspired artificial neural networks can be used to realize the efficient perception and parallel processing of unstructured data. This paper proposes a feedback‐controlled response system based on a NO2‐detecting artificial sensory synapse, which can process, judge, and react to a varying gas environment. The NO2‐detecting artificial sensory synapse adopts an organic heterostructure involving the charge trapping layer (pentacene) and hole‐conducting layer (copper‐phthalocyanine). The electron‐withdrawing nature of NO2 and its high compatibility with copper‐phthalocyanine induce the retentive behavior of an increase in the conductance at the hole conduction channel when consecutive positive pulses are applied to the gate terminal. The system consists of the artificial sensory synapse and artificial neuron circuits, which can provide systematic responses to varying NO2 conditions, thereby successfully simulating the efficient risk‐response system of biological neural networks. The proposed feedback‐controlled response system can facilitate the development of bionic electronics and artificial intelligence frameworks. [ABSTRACT FROM AUTHOR]

Approaching to the biological neural network, small-world neural networks have been demonstrated to improve the generalization performance of artificial neural networks. However, the architecture of small-world neural networks is typically large and predefined. This may cause the problems of overfitting and time consuming, and cannot obtain an optimal network structure automatically for a given problem. To solve the above problems, this paper proposes a pruning feedforward small-world neural network (PFSWNN), and applies it to nonlinear system modeling. Firstly, a feedforward small-world neural network (FSWNN) is constructed according to the rewiring rule of Watts–Strogatz. Secondly, the importance of each hidden neuron is evaluated based on its Katz centrality. If the Katz centrality of a hidden neuron is below the predefined threshold, this neuron is considered to be an unimportant node and then merged with its most correlated neuron in the same hidden layer. The connection weights are trained using the gradient-based algorithm, and the convergence of the proposed PFSWNN is theoretically analyzed in this paper. Finally, the PFSWNN model is tested on some problems for nonlinear system modeling, including the approximation for a rapidly changing function, CATS missing time-series prediction, four benchmark problems of UCI public datasets and a practical problem for wastewater treatment process. Experimental results demonstrate that PFSWNN exhibits superior generalization performance by small-world property as well as the pruning algorithm, and the training time of PFSWNN is shortened owning to a compact structure. • The importance of the hidden nodes is measured using Katz centrality. • The small-worldness property and pruning improve the generalization performance of PFSWNN. • PFSWNN can self-adjust to a compact structure by the pruning algorithm. • The convergence of PFSWNN can be guaranteed. [ABSTRACT FROM AUTHOR]

The nonlinear autoregressive network with exogenous inputs (NARX) is a recurrent dynamic neural architecture, which is commonly used for input–output modeling of nonlinear dynamical systems. Due to the lack of accurate mathematical model for the performance analysis on surge protective devices (SPDs), with the input of fast rising time electromagnetic pulse (FREMP), the NARX neural network is employed to predict the response characteristics of SPD in this paper. In order to verify the feasibility of this method, SPD test system is set up according to IEC 61000-4-24. The results show that the response curve estimated by the proposed model is in good agreement with experimental results, especially the waveform parameters such as response time, pulse peak, and residual voltage. The good forecasting performance of the network suggests that the NARX model used in this paper has good generalization ability. Moreover, with less measured data it can predict the response of the SPD under different voltage levels that were not yet measured. [ABSTRACT FROM AUTHOR]

This paper is concerned with quasi-synchronization of delayed memristive neural networks (MNNs) with switching jumps mismatches via aperiodically intermittent control. The issue is presented for three reasons: 1) the existing controllers for synchronization may be too complicated and not economical; 2) under the influence of switching jumps mismatches, synchronization of MNNs may fail to achieve; and 3) matrix measure method is less conservative but cannot be applied directly to synchronization of MNNs. To overcome these difficulties, the concept of asynchronously switching time interval is proposed to describe the phenomenon when the drive-response MNNs switch their connection weights asynchronously. Then, aperiodically intermittent control is designed and quasi-synchronization analysis is carried out based on a combined method that compromises the merits of interval matrix method and matrix measure method. A quasi-synchronization criterion, expressed in terms of the mixture of ${p}$ -norm and matrix measure of the memristive connection weights, is established. Meanwhile, the fundamental reason for the failure of complete synchronization is revealed. Moreover, an explicit expression of the error level is obtained and the design of the controller under a predetermined error level is presented. The obtained results in this paper reduce the conservativeness and provide a novel insight into the research of synchronization of MNNs. [ABSTRACT FROM AUTHOR]

Driven by the growing research enthusiasm for neuromorphic photonics, optical neurons mimicking biological properties based on microresonant cavities and lasers are attracting increasing attention, in which spiking synchronization of couple networks consisting of optical neurons based on lasers has been systematically investigated, and the emergence of neural network with biological properties based on microrings is foreseen. In this paper, considering the importance of oscillatory synchronization between neurons to achieve the basic biological functions, we investigate the period of self-pulsation which is fundamental oscillation within microring as a function of input light power and detuning, and achieve self-pulsation period synchronization of two mutually coupled microrings. In addition, the self-pulsation synchronization in time series between the unidirectionally coupled driving microring and responding microring is achieved by adding delayed feedback and adjusting key parameters, and this synchronization still exists even with the inclusion of more responding microrings. Our work provides basic investigations of the self-pulsation period as well as synchronization properties between coupled optical neurons based on microrings, and would be helpful for different concepts of neuromorphic computing based on microring, including artificial neural networks (ANN), spike neural networks (SNN) and reservoir computing. • The self-pulsation period of microring is investigated. • The self-pulsation period synchronization of microrings is achieved. • The self-pulsation synchronization of microrings in time series is achieved. • The expandability of self-pulsation synchronization in time series is discussed. • The investigation has a facilitating effect on neuromorphic computing. [ABSTRACT FROM AUTHOR]