Multi-Valued Neuron (MVN) was proposed for pattern classification. It operates with complex-valued inputs, outputs, and weights, and its learning algorithm is based on error-correcting rule. The activation function of...Multi-Valued Neuron (MVN) was proposed for pattern classification. It operates with complex-valued inputs, outputs, and weights, and its learning algorithm is based on error-correcting rule. The activation function of MVN is not differentiable. Therefore, we can not apply backpropagation when constructing multilayer structures. In this paper, we propose a new neuron model, MVN-sig, to simulate the mechanism of MVN with differentiable activation function. We expect MVN-sig to achieve higher performance than MVN. We run several classification benchmark datasets to compare the performance of MVN-sig with that of MVN. The experimental results show a good potential to develop a multilayer networks based on MVN-sig.展开更多
Carbon nanotubes have been used as scaffolds for tissue engineering. However, the identification of these nanomaterials disperses in biological solutions and their direct interaction with nerve cells is still controve...Carbon nanotubes have been used as scaffolds for tissue engineering. However, the identification of these nanomaterials disperses in biological solutions and their direct interaction with nerve cells is still controversial. The aim of this work is to characterize the topographic and ultra-structural distribution of the composite made of multi wall carbon nanotubes-polyvinylpyrrolidone (MWCNTs-PVP) deposited on the Helix aspersa neurons and over glass coverslip. Scanning Electron Microscopy (SEM) and Confocal Microscopy (CM) studies were done to analyze the properties of such MWCNTs-PVP composite. The cerebral ganglion of Helix aspersa was treated and incubated with MWCNTs-PVP, fixing it in paraformaldehyde at 4% and was observed with SEM and CM. Although the nanotubes were not labeled or stained with fluorescent compounds, the MWCNTs-PVP deposited on glass and on nerve cells, was observed by the confocal microscope in the reflection mode. In SEM studies, it was observed that MWCNTs-PVP was attached to the surface on neurons. Moreover, in CM studies, it was possible to observe that MWCNTs-PVP was attached to the neuronal membrane, crossing the cell membrane and getting into the cytoplasm. These results support the hypothesis that carbon nanotubes interact with the neuronal cell membrane and can be useful for neuronal tissue engineering. In addition, these results open new alternatives for toxicological studies, in order to elucidate the cytotoxicity of MWCNTs-PVP composite in neurons and other excitable cells.展开更多
由于传统的互补金属-氧化物-半导体(Complementary Metal Oxide Semiconductor,CMOS)神经元电路与生物学的契合性较差且电路复杂,提出了一种基于忆阻器的多端口输入的泄露-整合-激发(Leaky-Integrate-Fire,LIF)神经元电路。该电路由运...由于传统的互补金属-氧化物-半导体(Complementary Metal Oxide Semiconductor,CMOS)神经元电路与生物学的契合性较差且电路复杂,提出了一种基于忆阻器的多端口输入的泄露-整合-激发(Leaky-Integrate-Fire,LIF)神经元电路。该电路由运放、逻辑门等器件以及忆阻器构成,主要分为信号叠加模块和神经元信号产生模块。通过施加多个双尖峰脉冲信号并调节输入信号的数量和频率,模拟了生物神经元受到的不同程度刺激。研究发现施加到神经元上信号的数量和频率达到一定的值,神经元电路才会输出电压信号,这与生物体中只有受到一定程度的刺激时才会做出反应的现象是一致的。进一步,调节该电路中神经元信号产生模块的阈值电压大小,研究发现输入相同的信号,只有当电路的阈值电压较低时,神经元电路才能输出电压信号,这与生物中不同部位受到相同的刺激,神经元兴奋程度越高,越容易做出反应的现象一致。由此,该文所提出的LIF神经元电路不仅解决了传统电路输入信号单一、输入信号波形与生物信号波形差异大等问题,而且能模拟生物神经元的兴奋程度,这为人工神经网络的设计提供理论依据。展开更多
文摘Multi-Valued Neuron (MVN) was proposed for pattern classification. It operates with complex-valued inputs, outputs, and weights, and its learning algorithm is based on error-correcting rule. The activation function of MVN is not differentiable. Therefore, we can not apply backpropagation when constructing multilayer structures. In this paper, we propose a new neuron model, MVN-sig, to simulate the mechanism of MVN with differentiable activation function. We expect MVN-sig to achieve higher performance than MVN. We run several classification benchmark datasets to compare the performance of MVN-sig with that of MVN. The experimental results show a good potential to develop a multilayer networks based on MVN-sig.
文摘Carbon nanotubes have been used as scaffolds for tissue engineering. However, the identification of these nanomaterials disperses in biological solutions and their direct interaction with nerve cells is still controversial. The aim of this work is to characterize the topographic and ultra-structural distribution of the composite made of multi wall carbon nanotubes-polyvinylpyrrolidone (MWCNTs-PVP) deposited on the Helix aspersa neurons and over glass coverslip. Scanning Electron Microscopy (SEM) and Confocal Microscopy (CM) studies were done to analyze the properties of such MWCNTs-PVP composite. The cerebral ganglion of Helix aspersa was treated and incubated with MWCNTs-PVP, fixing it in paraformaldehyde at 4% and was observed with SEM and CM. Although the nanotubes were not labeled or stained with fluorescent compounds, the MWCNTs-PVP deposited on glass and on nerve cells, was observed by the confocal microscope in the reflection mode. In SEM studies, it was observed that MWCNTs-PVP was attached to the surface on neurons. Moreover, in CM studies, it was possible to observe that MWCNTs-PVP was attached to the neuronal membrane, crossing the cell membrane and getting into the cytoplasm. These results support the hypothesis that carbon nanotubes interact with the neuronal cell membrane and can be useful for neuronal tissue engineering. In addition, these results open new alternatives for toxicological studies, in order to elucidate the cytotoxicity of MWCNTs-PVP composite in neurons and other excitable cells.
文摘由于传统的互补金属-氧化物-半导体(Complementary Metal Oxide Semiconductor,CMOS)神经元电路与生物学的契合性较差且电路复杂,提出了一种基于忆阻器的多端口输入的泄露-整合-激发(Leaky-Integrate-Fire,LIF)神经元电路。该电路由运放、逻辑门等器件以及忆阻器构成,主要分为信号叠加模块和神经元信号产生模块。通过施加多个双尖峰脉冲信号并调节输入信号的数量和频率,模拟了生物神经元受到的不同程度刺激。研究发现施加到神经元上信号的数量和频率达到一定的值,神经元电路才会输出电压信号,这与生物体中只有受到一定程度的刺激时才会做出反应的现象是一致的。进一步,调节该电路中神经元信号产生模块的阈值电压大小,研究发现输入相同的信号,只有当电路的阈值电压较低时,神经元电路才能输出电压信号,这与生物中不同部位受到相同的刺激,神经元兴奋程度越高,越容易做出反应的现象一致。由此,该文所提出的LIF神经元电路不仅解决了传统电路输入信号单一、输入信号波形与生物信号波形差异大等问题,而且能模拟生物神经元的兴奋程度,这为人工神经网络的设计提供理论依据。
