| 1. | Importance of optimizing neural activation function types
优化神经元激活函数类型的重要性 |
| 2. | We analyze global stability of a class of neural networks with discrete delays under three assumptions of activation functions
在对神经网络的激励函数的三个假设下，研究了具有离散时滞的神经网络的稳定性。 |
| 3. | Based on analyzing the multilayer perceptron neural network , it induces the error backpropagation algorithm while taking the hyperboloid tangential function as nonlinear activation function
在分析多层感知器神经网络的基础上，推导了双曲正切函数为激励函数的误差后向传播算法。 |
| 4. | Second by using liapunov direct method , two different sufficient global exponential stability conditions are obtained on condition that activation functions are lipschitz continuous
第二种是采用李雅普诺夫直接方法，在激励函数的lipschitz连续假设下，得到了两个全局指数稳定的充分性判据。 |
| 5. | The lumping scheme and reaction network of this model largely decrease the complexity of model and demand for detailed component analysis , so it is easy to apply this model to the commercial units . the catalyst coking model with average activation function of catalyst is presented
深入探讨了重整催化剂的结焦失活机理，首次提出了连续重整装置催化剂失活的平均活性函数概念，给出了催化剂结焦失活模型。 |
| 6. | And they can run efficiently even if the data are incomplete and noisy . so we advance the neural fuzzy cooperate system - combined activation function ( nfcs - cap ) algorithms . it fuses strongpoint of both neural net and fuzzy system and adept to mine rules in database especially
而模糊神经网络由于其本身良好的鲁棒性、自组织自适应性、并行处理、分布存储、高度容错、易于理解等特性，特别适用于在以上情况下用来挖掘数据库中的知识。 |
| 7. | Wavelet analysis and neural network have been researched . based on bp neural network , analysing the features of the bp neural network ' s activation function and the structure of the bp neural network , wavelets neural network is constructed with wavelet function . fourier transform has a important role on signal processing , so the fourier transform is introduced . a function ' fourier transform is analysed . after introducing the concept of the wavelet analysis , a serval wavelet functions is given out , so the knowledge of wavelet analysis is better understood . the multiresolution analysis has a important role on constructing wavelet function and understanding the wavelet transform
小波分析和神经网络得到了广泛的研究，小波神经网络是在bp神经网络的基础上，考虑和分析了bp神经网络的激励函数的特点，以及bp神经网络的结构，结合了小波分析的知识而构造的。傅里叶变换在信号的处理起重要的作用，所以介绍了傅里叶变换、离散傅里叶变换和加窗傅里叶变换，并对一函数作离散傅里叶变换分析。 |
| 8. | The commonly seen gold - carrying minerals in hard - to - leach gold ores - pyrite and arsonpyrite were used in the experiment . through sweep electron microscope ( sem ) and energy dispersive x - ray spectrometer ( edx ) , the change law of the shape and appearance characteristics of the gold - carrying minerals under the mechanical activation function of drum grinding and other activation function were studied so as to provide a reliable basis for the study on the mechanism of the effect of mechanical activiation on the intensification of leaching process
采用难浸金矿常见的载金矿物? ?黄铁矿和毒砂做试验，通过扫描电镜( sem )和x射线能谱仪( edx )研究了在滚筒磨的机械活化作用和不同的活化条件下载金矿物形貌特征的变化规律，为弄清机械活化作用强化浸出过程的机理提供了可靠的研究依据。 |
| 9. | The dissertation is dedicatd to the theory of maximum nongaussianity estimation in ica , and some problems are emphasized , e . g . , the uniqueness of estimation , the convergence of fastica , the constraint of independence and the nonlinear activation function , the algorithms for multiple components and the order of independnent components
本文重点研究最大非高斯估计的相关理论问题,包括估计的唯一性,算法的收敛性,独立性约束分析和非线性激活函数,独立分量的排序和子空间的选择,具有时间结构信息的信号源的估计等。 |
| 10. | Artificial neural network ( ann ) is non - linear technology developed since 1980 . the ann model is determined by topology structure , activation function and learning method . the model possesses with good fault tolerance , self - adaptability and non - linear mapping , especially it is suitable to resolve complex causal problems
人工神经网络模型具有很强的容错性、自适应性和非线性的映射能力，特别适于解决因果关系复杂的非确定性推理、判断、识别和分类等问题。 |
