finalA05_exo3

一本讲述时间序列分析的英文书，非常详细！(A book about time series analysis of the English book, very detailed!)

采用PCA提取人脸特征,先计算M个图像的平均值脸，再求出各特征值脸。(Using PCA for face feature extraction, first calculate the average face image of M, and then find the face value of each characteristic.)

Approximation entropy: Approximation entropy which is used in time series to measure the irregularity

代码1用于将位置经纬高信息转换为地心地固坐标系下xyz信息；代码2是捷联惯导算法代码，用于将加速度计和陀螺仪输出的信息进行解算得到位置、速度和姿态信息。(Code 1 is used to convert the location of warp and weft information into XYZ information in a geocentric coordinate system; code 2 is a strapdown inertial navigation algorithm code, which is used to compute the information of accelerometers and gyroscopes to get position, speed and attitude information.)

qpsk通过信道信噪比估计，包括qpsk，8psk，16psk，16qam(qpsk through the channel SNR estimation)

说明：  粒子群算法（PSO）整定pid控制参数，比传统Z-N整定方法要好，内附matlab程序与simulink模型(Particle swarm optimization (PSO) is better than traditional Z-N tuning method in tuning PID control parameters. It includes matlab program and Simulink model.)

利用matlab实现了对于通话排队论阻塞率的仿真(Simulation on the blocking rate)

0_Linear：线性回归练习 1_LogistRegression：逻辑回归练习 2_boston：波士顿房价预测 3_regularization：正则化 5_stock：股票价格预测 6_CTR：点击率预测 7_all_model：回归与分类模型 8_muticlass：多分类练习 9_regression_implement：不同回归算法tf实现(0_Linear. 1_LogistRegression. 2_boston. 3_regularization. 5_stock. 6_CTR. 7_all_model. 8_muticlass. 9_regression_implement.)

一维时间序列分行维数，利用计盒法计算得到的分形维数，希望对大家有所帮助！(Branch of one-dimensional time series dimension, using box method calculated the fractal dimension, and they hope to be helpful to everyone!)

TVL1重建从混合压缩采样，rice大学泊松分布图像重建与全变分正规化的代码(Note: (1)Fore small-scale data, hybrid sampling consists of Bernoulli random sampling and low-resolution sampling (TVL1_HCS) (1)Fore large-scale data, hybrid sampling consists of Fourier radial sampling and low-frequency sampling (TV_RecPF) Copyright (C) 2010 Xianbiao Shu, Computer Vision and Robotics Lab, University of Illinois at Urbana-Champaign. Acknowledgement: RecPF (Rice University) PDCO (Stanford University) Twister (MIT) Reference: X. Shu, N. Ahuja, Hybrid Compressive Sampling via a New Total Variation TVL1, In Proceedings of ECCV 2010. Usage: >> run_me_first >>[Im_rec, PSNR_est1,PSNR_est2] = TVL1L2_HCS_test( Ball1_32.bmp , 0.35,3) >>[Im_rec, PSNR_est1,PSNR_est2] = TVL1L2_HCS_test( Ball1_32.bmp , 0.35,3) >>[Im_rec, PSNR1, PSNR2, G_error, Ratio] = TV_RecPF_test( Chest.png , 5, 10, 0, 1e-10, 0) )