• Title/Summary/Keyword: synaptic weight

Search Result 21, Processing Time 0.027 seconds

Implementation of Neuromorphic System with Si-based Floating-body Synaptic Transistors

  • Park, Jungjin;Kim, Hyungjin;Kwon, Min-Woo;Hwang, Sungmin;Baek, Myung-Hyun;Lee, Jeong-Jun;Jang, Taejin;Park, Byung-Gook
    • JSTS:Journal of Semiconductor Technology and Science
    • /
    • v.17 no.2
    • /
    • pp.210-215
    • /
    • 2017
  • We have developed the neuromorphic system that can work with the four-terminal Si-based synaptic devices and verified the operation of the system using simulation tool and printed-circuit-board (PCB). The symmetrical current mirrors connected to the n-channel and p-channel synaptic devices constitute the synaptic integration part to express the excitation and the inhibition mechanism of neurons, respectively. The number and the weight of the synaptic devices affect the amount of the current reproduced from the current mirror. The double-stage inverters controlling delay time and the NMOS with large threshold voltage ($V_T$) constitute the action-potential generation part. The generated action-potential is transmitted to next neuron and simultaneously returned to the back gate of the synaptic device for changing its weight based on spike-timing-dependent-plasticity (STDP).

Learning Model and Application of New Preceding Layer Driven MLP Neural Network (새로운 Preceding Layer Driven MLP 신경회로망의 학습 모델과 그 응용)

  • 한효진;김동훈;정호선
    • Journal of the Korean Institute of Telematics and Electronics B
    • /
    • v.28B no.12
    • /
    • pp.27-37
    • /
    • 1991
  • In this paper, the novel PLD (Preceding Layer Driven) MLP (Multi Layer Perceptron) neural network model and its learning algorithm is described. This learning algorithm is different from the conventional. This integer weights and hard limit function are used for synaptic weight values and activation function, respectively. The entire learning process is performed by layer-by-layer method. the number of layers can be varied with difficulty of training data. Since the synaptic weight values are integers, the synapse circuit can be easily implemented with CMOS. PLD MLP neural network was applied to English Characters, arbitrary waveform generation and spiral problem.

  • PDF

Improvement of Initial Weight Dependency of the Neural Network Model for Determination of Preconsolidation Pressure from Piezocone Test Result (피에조콘을 이용한 선행압밀하중 결정 신경망 모델의 초기 연결강도 의존성 개선)

  • Park, Sol-Ji;Joo, No-Ah;Park, Hyun-Il;Kim, Young-Sang
    • Proceedings of the Korean Geotechical Society Conference
    • /
    • 2009.03a
    • /
    • pp.456-463
    • /
    • 2009
  • The preconsolidation pressure has been commonly determined by oedometer test. However, it can also be determined by in-situ test, such as piezocone test with theoretical and(or) empirical correlations. Recently, Neural Network(NN) theory was applied and some models were proposed to estimate the preconsolidation pressure or OCR. However, since the optimization process of synaptic weights of NN model is dependent on the initial synaptic weights, NN models which are trained with different initial weights can't avoid the variability on prediction result for new database even though they have same structure and use same transfer function. In this study, Committee Neural Network(CNN) model is proposed to improve the initial weight dependency of multi-layered neural network model on the prediction of preconsolidation pressure of soft clay from piezocone test result. It was found that even though the NN model has the optimized structure for given training data set, it still has the initial weight dependency, while the proposed CNN model can improve the initial weight dependency of the NN model and provide a consistent and precise inference result than existing NN models.

  • PDF

CMOS Analog Integrate-and-fire Neuron Circuit for Driving Memristor based on RRAM

  • Kwon, Min-Woo;Baek, Myung-Hyun;Park, Jungjin;Kim, Hyungjin;Hwang, Sungmin;Park, Byung-Gook
    • JSTS:Journal of Semiconductor Technology and Science
    • /
    • v.17 no.2
    • /
    • pp.174-179
    • /
    • 2017
  • We designed the CMOS analog integrate and fire (I&F) neuron circuit for driving memristor based on resistive-switching random access memory (RRAM). And we fabricated the RRAM device that have $HfO_2$ switching layer using atomic layer deposition (ALD). The RRAM device has gradual set and reset characteristics. By spice modeling of the synaptic device, we performed circuit simulation of synaptic device and CMOS neuron circuit. The neuron circuit consists of a current mirror for spatial integration, a capacitor for temporal integration, two inverters for pulse generation, a refractory part, and finally a feedback part for learning of the RRAM. We emulated the spike-timing-dependent-plasticity (STDP) characteristic that is performed automatically by pre-synaptic pulse and feedback signal of the neuron circuit. By STDP characteristics, the synaptic weight, conductance of the RRAM, is changed without additional control circuit.

A Study on the Synaptic Characteristics of SONOS memories for the Artificial Neural Networks (인공신경망을 위한 SONOS 기억소자의 시냅스특성에 관한 연구)

  • 이성배;김주연;서광열
    • Journal of the Korean Institute of Electrical and Electronic Material Engineers
    • /
    • v.11 no.1
    • /
    • pp.7-11
    • /
    • 1998
  • In this paper, a new synapse cell with nonvolatile SONOS semiconductor memory device is proposed and it's fundamental function electronically implemented SONOS NVSM has shown characteristics that the memory value, synaptic weights, can be increased or decreased incrementally. A novel SONOS synapse is used to read out the stored analog value. For the purpose of synapse implementation using SONOS NVSM, this work has investigated multiplying characteristics including weight updating characteristics and neuron output characteristics. It is concluded that SONOS synapse cell has good agreement for use as a synapse in artificial neural networks.

  • PDF

The Effects of Complex Motor Training on Motor Function and Synaptic Plasticity After Neonatal Binge-like Alcohol Exposure in Rats (복합운동훈련이 신생 흰쥐의 알코올성 소뇌손상 후 운동기능 및 신경연접가소성에 미치는 영향)

  • Lee, Sun-Min;Koo, Hyun-Mo;Kwon, Hyuk-Cheol
    • Physical Therapy Korea
    • /
    • v.12 no.3
    • /
    • pp.56-66
    • /
    • 2005
  • The purposes of this study were to test that complex motor training enhance motor function significantly, to test change in cerebellum, and to test the synaptic plasticity into the immunohistochemistry response of synaptophysin. Using an animal model of fetal alcohol syndrome - which equates peak blood alcohol concentrations across developmental period - the effects of alcohol on body weight during periods were examined. The effect of complex motor training on motor function and synaptic plasticity of rat exposed alcohol on postnatal days 4 through 10 were studied. Newborn rats were assigned to one of two groups: (1) normal group (NG), via artificial rearing to milk formula and (2) alcohol groups (AG), via 4.5 g/kg/day of ethanol in a milk solution. After completion of the treatments, the pups were fostered back to lactating dams, where they were raised in standard cages (two-and three animals per cage) until they were postnatal 48 days. Rats from alcohol group of postnatal treatment then spent 10 days in one of two groups: Alcohol-experimental group was had got complex motor training (learning traverse a set of 6 elevated obstacles) for 4 weeks. The alcohol-control group was not trained. Before consider replacing with "the experiment/study", (avoid using "got" in writing) the rats were examined during four behavioral tests and their body weights were measured, then their coronal sections were processed in rabbit polyclonal antibody synaptophysin. The synaptophysin expression in the cerebellar cortex was investigated using a light microscope. The results of this study were as follows: 1. The alcohol groups contained significantly higher alcohol concentrations than the normal group. 2. The alcohol groups had significantly lower body weights than the normal group. 3. In alcohol groups performed significantly lower than the normal group on the motor behavioral test. 4. In alcohol-control group showed significantly decreased immunohistochemistric response of the synaptophysin in the cerebellar cortex compared to the nomal group. These results suggest that improved motor function induced by complex motor training after postnatal exposure is associated with dynamically altered expression of synaptophysin in cerebellar cortex and that is related with synaptic plasticity. Also, these data can potentially serve as a model for therapeutic intervention.

  • PDF

Acute and Subacute Effect of Lead acetate on Enzyme Activities and Ultrastructure in Mouse Diencephalone (초산납이 생쥐 간뇌의 미세구조 및 Catecholamine 대사에 미치는 영향)

  • Lee, Jung-Hee;Yoo, Chang-Kyu;Choe, Rim-Soon
    • Applied Microscopy
    • /
    • v.18 no.2
    • /
    • pp.187-204
    • /
    • 1988
  • The present experiment was performed to investigate the acute and subacute effect of lead acetate on ultrastructural and biochemical changes in mouse diencephalon. In acute case, mouse were peritoneally injected with lead acetate at a dose of 0.26 mmole/kg body weight, and after treatment, mouse were sacrificed at time intervals of 12, 24, 48, and 96 hours. In subacute case, mouse were injected at doses of 0.07 mmoie/kg B. W. and 0.13 mmole/kg B.W. once at two days, and after treatment, mouse wee sacrificed at 1 week, 2 weeks, and 3 weeks. It was observed that after acute treatment, changes composed of increased monoamine oxidase activity, $Na^{+}-K^{+}$ ATPase activity, decreased $Mg^{2+}$-APTase activity, wrinkled myelin, swollen Golgi apparatus and more dense synaptic vesicle in nerve terminal. After subacute treatment, decreased monoamine oxidase activity, increased $Mg^{2+}$-ATPase, $Na^{+}-K^{+}$ ATPase, lose of myelin, uneven mitochondrial distribution, synaptic vesicular density and edema, but at a higher dose the effect was more severe. Therefore, lead acetate caused abnormal change of diencephalon, and at a subacute, it appears metal accumulative toxicity.

  • PDF

Time Series Prediction Using a Multi-layer Neural Network with Low Pass Filter Characteristics (저주파 필터 특성을 갖는 다층 구조 신경망을 이용한 시계열 데이터 예측)

  • Min-Ho Lee
    • Journal of Advanced Marine Engineering and Technology
    • /
    • v.21 no.1
    • /
    • pp.66-70
    • /
    • 1997
  • In this paper a new learning algorithm for curvature smoothing and improved generalization for multi-layer neural networks is proposed. To enhance the generalization ability a constraint term of hidden neuron activations is added to the conventional output error, which gives the curvature smoothing characteristics to multi-layer neural networks. When the total cost consisted of the output error and hidden error is minimized by gradient-descent methods, the additional descent term gives not only the Hebbian learning but also the synaptic weight decay. Therefore it incorporates error back-propagation, Hebbian, and weight decay, and additional computational requirements to the standard error back-propagation is negligible. From the computer simulation of the time series prediction with Santafe competition data it is shown that the proposed learning algorithm gives much better generalization performance.

  • PDF

Structural Optimization and Improvement of Initial Weight Dependency of the Neural Network Model for Determination of Preconsolidation Pressure from Piezocone Test Result (피에조콘을 이용한 선행압밀하중 결정 신경망 모델의 구조 최적화 및 초기 연결강도 의존성 개선)

  • Kim, Young-Sang;Joo, No-Ah;Park, Hyun-Il;Park, Sol-Ji
    • KSCE Journal of Civil and Environmental Engineering Research
    • /
    • v.29 no.3C
    • /
    • pp.115-125
    • /
    • 2009
  • The preconsolidation pressure has been commonly determined by oedometer test. However, it can also be determined by insitu test, such as piezocone test with theoretical and(or) empirical correlations. Recently, Neural Network (NN) theory was applied and some models were proposed to estimate the preconsolidation pressure or OCR. It was already found that NN model can come over the site dependency and prediction accuracy is greatly improved when compared with present theoretical and empirical models. However, since the optimization process of synaptic weights of NN model is dependent on the initial synaptic weights, NN models which are trained with different initial weights can't avoid the variability on prediction result for new database even though they have same structure and use same transfer function. In this study, Committee Neural Network (CNN) model is proposed to improve the initial weight dependency of multi-layered neural network model on the prediction of preconsolidation pressure of soft clay from piezocone test result. Prediction results of CNN model are compared with those of conventional empirical and theoretical models and multi-layered neural network model, which has the optimized structure. It was found that even though the NN model has the optimized structure for given training data set, it still has the initial weight dependency, while the proposed CNN model can improve the initial weight dependency of the NN model and provide a consistent and precise inference result than existing NN models.

Exercise alleviates cisplatin-induced toxicity in the hippocampus of mice by inhibiting neuroinflammation and improving synaptic plasticity

  • Se Hwan Park;Jeong Rim Ko;Jin Han
    • The Korean Journal of Physiology and Pharmacology
    • /
    • v.28 no.2
    • /
    • pp.145-152
    • /
    • 2024
  • Chemotherapy-induced cognitive impairment is recognized as the most typical symptom in patients with cancer that occurs during and following the chemotherapy treatment. Recently many studies focused on pharmaceutical strategies to control the chemotherapy side effects, however it is far from satisfactory. There may be a need for more effective treatment options. The aim of this study was to investigate the protective effect of exercise on cisplatin-induced neurotoxicity. Eight-week-old C57BL6 mice were separated into three group: normal control (CON, n = 8); cisplatin injection control (Cis-CON, n = 8); cisplatin with aerobic exercise (Cis-EXE, n = 8). Cisplatin was administered intraperitoneally at a dose of 3.5 mg/kg/day. The Cis-EXE group exercise by treadmill running (14-16 m/min for 45 min daily, 3 times/week) for 12 weeks. Compared to the CON group, the cisplatin injection groups showed significant decrease in body weight and food intake, indicating successful induction of cisplatin toxicity. The Cis-CON group showed significantly increased levels of pro-inflammatory cytokines including IL-6, IL-1β, and TNF-α in the hippocampus, while the Cis-EXE group was significantly decreased in the expression of IL-6, IL-1β, and TNF-α. In addition, compared to the CON group, the levels of synapse-related proteins including synapsin-1 and -2 were significantly reduced in the Cis-CON group, and there was a significant difference between the Cis-CON and Cis-EXE groups. Antioxidant and apoptosis factors were significantly improved in the Cis-EXE group compared with the Cis-CON group. This study suggest that exercise could be meaningful approach to prevent or improve cisplatin-induced cognitive impairment.