suanfaLyapunov

这是 Andrew Ng 机器学习讲义《Reinforcement Learning and Control》,包含了离散和连续MDP的内容，比官网的课件全。（官网的缺乏连续MDP部分）(This is Andrew Ng Machine Learning Materials " Reinforcement Learning and Control" , contains a discrete and continuous MDP content than the official website of the courseware whole. (Official website of the lack of continuous MDP section))

支持向量机的多分类问题，主要是将其分成了5类(Support Vector Machine)

《精通MATLAB综合辅导与指南》里面的例程(" Comprehensive Counseling and proficient in MATLAB Guide" inside the routine)

DualThurst策略，经过改编，盈利能力较好(The DualThurst strategy has been adapted better profitability)

视频读取及目标背景提取，matlab平台下可以对视频取帧(Read and goals video background extraction, matlab platform can take video frame)

搭建simulink仿真框图，使用sisotool工具箱对对象传递函数进行PID参数整定。由上面所得到的PID控制器中PD部分的控制经验可获取模糊控制器的控制规则，设计Mamdany型模糊控制器和T-S型模糊控制器。 (Build a simulink simulation diagram, use sisotool toolbox object transfer function PID parameter tuning. PD control experience PID controller portion obtained by the above control rules can get fuzzy controller design Mamdany TS fuzzy controller and fuzzy controller.)

说明：  用于检测图像的复制粘贴，包括旋转、缩放等，算法是基于sift(Copy and paste for detecting images)

Simulation PHY layer of Wimax 802.16e, including Tx, Channel and Rx, baseband only.

说明：  蒙特卡洛仿真对二元antipodal通信系统的误比特率的估计。(Monte Carlo simulation of binary antipodal communication system bit error rate estimate.)

Spline2D is a work in progress. Currently it only matches C0 (continuous function) not C1 (continuous in 1st derivative) or C2 (continuous in 2nd derivative), which will be added later.