对于初学者具有参考意义，基于分段非线性权重值的Pso算法，包括主成分分析、因子分析、贝叶斯分析，自己编的5种调制信号，Relief计算分类权重，包含特征值与特征向量的提取、训练样本以及最后的识别，是学习PCA特征提取的很好的学习资料。- For beginners with a reference value, Based on piecewise nonlinear weight value Pso algorithm, Including principal component analysis, factor analysis, Bayesian analysis, Own five modulation signal, Relief computing classification weight, Contains the eigenvalue and eigenvector extraction, the training sample, and the final recognition, Is a good learning materials to learn PCA feature extraction.