qpsk

说明：  BP算法是应用最广泛的神经网络算法，常用于预测，得到了很好的效果(BP algorithm is the most widely used neural network algorithm, commonly used in forecasting, we get good results)

这是QAM信号和PSK信号产生程序，可以生成16-QAM,32-QAM,64-QAM和M进制psk信号。(This is PSK and QAM signals a process can generate 16-QAM ,32-QAM. 64-M and 229 QAM signals psk.)

详细介绍了机器学习中非监督学习的内容、分类和方法(This document introduces the unsupervised learing of Machine learning with details.)

说明：  第四章 MATLAB 图形对象操作 4.1 图形对象的属性 4.2 图形对象句柄的获取 4.2.1 对象创建时获取 4.2.2 层次关系来获取 4.2.3 当前对象的获取 4.2.4 根据对象属性值的获取 4.3 图形对象句柄的删除与判断 4.3.1 句柄的删除 4.3.2 句柄的判断 4.4 图形对象属性值的获取与设置 4.4.1 图形对象属性值的设置 4.4.2 图形对象属性值的获取 4.4.3 用户缺省值的操作 4.5 图形对象的其它操作 4.5.1 figflag 函数 4.5.2 findfigs 函数 4.5.3 copyobj 函数 4.5.4 capture 函数 4.5.5 refresh 函数 4.5.6 saveas 函数 4.5.7 hgload 函数和 hgsave 函数 4.5.8 newplot 函数 (Chapter IV of MATLAB graphic objects to operate the properties of graphic object 4.1 graphic object handle 4.2 Access 4.2.1 object created when the relationship between the level of access to 4.2.2 to get 4.2.3 to obtain the current object under the object attribute value 4.2.4 Access 4.3 graphic object handle to remove the judge handles the deletion of 4.3.1 judgments 4.3.2 Handle 4.4 graphic object attribute values to obtain and set the target attribute value 4.4.1 graphics settings 4.4.2 graphic object attribute values to obtain 4.4 .3 user operation 4.5 Default graphic object 4.5.1 figflag other operations function 4.5.2 findfigs function 4.5.3 copyobj function 4.5.4 capture function 4.5.5 refresh function 4.5.6 saveas function 4.5.7 hgload function and 4.5.8 newplot function hgsave function)

说明：  包括二分法，Newton下山法和improved Newton迭代法(Including the dichotomy, Newton downhill method and improved Newton iteration method)

matlab矩阵操作大全，还包括一些特殊矩阵的生成(matlab matrix operations Daquan, including the formation of some special matrices)

Forward Backward Propogation in Information Form

这是一个支持向量机回归预测的代码，适合于初学者来体验svr的回归能力( U8FD9 u662F u4E00 u4E2A u652F u6301 u5411 u91CF u673A u56DE u5F52 u9884 u6D4B u7684 u4EE3 u7801 uFF0C u9002 u5408 u4E8E u521D u5B66 u8005 u6765 u4F53 U9A8Csvr u7684 u56DE u5F52 u80FD u529B)

平滑IMM算法,用于机动目标的跟踪,并与常用的IMM滤波估计算法进行了比较(IMM smoothing algorithm for maneuvering target tracking, and with commonly used filtering IMM estimation algorithm are compared)

神经网络用于分类与回归,先产生训练样本与测试样本，每一列为一个样本( neural network for classification and regression, first create the training samples and test samples, each as a sample of)