NN_DESIGN(code)

MT8870 sample code In this example you learn how to work with MT8870.

Enhancement of LEACH

由一个隐马尔可夫模型产生一个观测值序列和一个状态序列(Consists of a hidden Markov model to generate a sequence of observations and a state sequence)

说明：  小波神经网络 基于交通流量的预测 具有很强的模型转换能力(Wavelet neural network traffic prediction based on model transformation has a strong ability to)

基于Elman神经网络的电力负荷预测模型研究(Research on power load forecasting model based on Elman neural network)

使用gevp解线性矩阵不等式的程序，很完整，有注释，对于学习求解线性矩阵不等式很有帮助(Solving linear matrix inequalities using gevp procedures are complete, with annotations, for learning useful for solving linear matrix inequalities)

基于Alamouti的8PSK调制2发1收程序。(8PSK modulation based on Alamouti' s 2 made a settlement process.)

斯坦福大学机器学习课程的一个源代码例子，对于机器学习和deep learning的学习研究很有帮助(this codes are from standford university,which are about machine learning and deep learning.they will be helpful in this research field)

对回波信号实现仿真 及其不同方法的对比 对线性调频信号的仿真等。(Realize the simulation of linear frequency modulation signal compression And the echo signal simulation, comparison of different methods of implementation. Listen Read phonetically Dictionary- View detailed dictionary )

把HHT和神经网络结合起来，识别出是否故障。对于HHT变换，这里用到的是它的IMF分解，然后用能量理论来判别哪些模组是虚假分量，哪些是是真实分量。对于EMD分解后的谱进行特征提取 ，利用的理论基础就是模糊熵。计算出真实 分量的模糊熵，作为输入层；输出层为两个神经元，为0（故障 ）1（正常 （正常 ）判别该信号是否出现故障。(The combination of HHT and neural network to identify whether the fault. For the HHT transform, here is the use of its IMF decomposition, and then use the energy theory to determine which module is false, which is true component. The theoretical basis for the feature extraction of the spectrum after EMD decomposition is fuzzy entropy. The fuzzy entropy of the real component is calculated, as the input layer, the output layer is two neurons, which is 0 (fault) 1 (normal) to determine whether the signal is faulty.)