SpaceLib_MatLab_manual

VC++实现的离散余弦变换，用于matlab环境，与系统提供的不同。(VC++ implementation of the discrete cosine transform for the matlab environment, and the system provides different.)

It is used for input

Learning_random.m : 随机选择样例，从（90）pool里随机选择样本，删除版本空间树的数量 activeLearning.m:根据最大熵原则,从pool里选择样本，删除版本空间树的数量 getlabel.m:根据树和测试样例，得到根据该树得到的类标 getTrees.m:从提供的大量树结构(随机生成的)中，随机挑选一定数量的树(如果该树的叶子节点无类标随机添加) RandomCreateTree_d_unbalance:根据样本点割点中的非平衡割点建造树， RandomCreateTree_d_all.m:根据所有样本点的割点建造树 randomdata.m：给定属性取值，造数据 randomselect.m:从数据中随机选择一部分作为 showTree.m:显示树的结构 test.m:给出树，测试样例，得到正确率(Learning_random.m: randomly selected sample, randomly selected sample from (90) pool the The deleted version space tree quantity activeLearning.m: selecting a sample from the pool based on the principle of maximum entropy, delete the number of version space tree getlabel.m: According to the tree and the test sample obtained according to the class standard getTrees.m the tree: from the tree structure (randomly generated), randomly selected a certain number of trees (the leaves of the tree node class marked randomly adding ) RandomCreateTree_d_unbalance: According to the sample point cut point unbalanced cut point construction tree, RandomCreateTree_d_all.m: construction of the tree randomdata.m all sample points cut point: given the value of the property, manufacturing data randomselect.m: random data Select as part showTree.m: tree structure test.m: tree, the test sample is given to get the correct rate)

实现多模态过程监测功能,包括数据的预处理，标准化处理等(multimode process detection)

说明：  基于卡尔曼滤波器的PID控制的matlab程序。(Kalman filter-based PID control)

Sample Matrix Inversion is implemented via matlab. the code plot the SINR performance as a function of BER and compare it to the theoretical upper bound BER. YOU CAN CHANGE THE PARAMETERS AND TEST THE RESULTS.

说明：  ∑-Δ调制工具箱2.0版做∑-Δ调制用仿真分析(Σ-Δ Modulation Toolkit 2.0)

Digital Modulations msk gmsk

two-way-relay场景下，中继处理矩阵的设计，仿真分析了DPC-ZF及DPC-MMSE算法的性能要由于AF与ZF算法性能。(In the two-way-relay scenario,simulation show that both DPC-ZF and DPC-MMSE have a higher throughput and lower BER performance compared with the traditional AF algorithm and ZF algorithm.)

As significant work was done in machine learning in past two decades, we have limited to recent articles in machine learning. A brief review of machine learning and its application can found in [Duttion and Conroy,1996). The most common supervised learning method and major implementation to various machine learning applications is artificial neural network as ANNs.