ICA-algorithm-and-articles

This program computes the 2D IDCT as per the equation. The result matches the output of the built-in Matlab function.Good for analysis.

用有效折射率法计算长周期光纤光栅的透射光谱情况，即是通过有效折射率算法，找出损耗峰（共振峰）。(Simulate transmission spectrum of LPFG via using effect refractive index algorithm. )

该函数实现扩展卡尔曼滤波算法，用于非线性状态估计，目标跟踪(EKF filter)

将含有高斯白噪声的输入信号通过线性系统后（带阻滤波器）的输入、输出前信号的频谱、功率谱对比。同时给出了线性系统特性(Gaussian white noise, containing the input signal by a linear system (bandstop filter) of the input, before the output signal spectrum, the power spectrum of contrast. Also gives a linear system characteristics)

BP神经网络公路运量预测 实例讲述如何在matlab中使用BP神经网络(BP neural network prediction road traffic)

TLD算法的QT版本，没有matlab的相关代码，完全用C/C++实现，可以移植到嵌入式平台或者其他平台(TLD algorithm for the QT version, no matlab related code, complete with C/C++ implementation can be transplanted to the embedded platform or other platforms)

驱动力和制动力是车辆纵向力学特性中重要部分，本程序采用matlab，根据发动机的万有特性曲线和循环工况，绘制出车辆的驱动力和制动力随时间的关系(The driving force and braking force is an important part of the vehicle longitudinal mechanical properties, the program uses matlab, according to the universal characteristic curve of the engine and the driving cycle, to map out the driving force and braking force of the vehicle over time)

在多径信号下的采用toeplitz来进行矩阵重构，对信号到达角进行估计(Using Toeplitz matrix reconstruction, DOA estimation under multipath signals )

说明：  数学形态滤波器用于旋转机械的振动信号的降噪(Morphological noise filter for rotating machinery vibration signals)

【转】协同过滤代码，用于推荐系统，包括基于项目和基于用户两种情况。实现基于用户和基于项目的协同过滤算法。 实验所用数据为MovieLens – a web-based movies recommender system with 43,000 users & over 3500 movies. 保存在ga.mat文件用，由于ga.test测试集过于庞大，全部用来计算的话耗时庞大，所以每次计算时随机选择部分，具体函数的使用请参照probar.m。我所得到的实验结果保存在results1-results8.mat里面。([Turn] collaborative filtering code for the recommended system, including project-based and user-based two cases. User-based and project-based collaborative filtering algorithm. Experimental data for MovieLens- a web-based movies recommender system with 43,000 users & over 3500 movies. Saved in ga.mat file with Because ga.test test set is too large, all used to calculate the time-consuming massive, so every time The calculated random selection portion, the use of the specific function, please refer to probar.m. I obtained experimental results are saved inside results1-results8.mat.)