deblock

I am submitting my CSE assignment no 4 which was based on Position square specific matrix (PSSM). Please find the attached zip file which contains matlab code of prob1 as CSEAssignment4_0800229.m file, report of the problem as Report.pdf and the two .pep files required for program to run.

Q 学习方面的MATLAB程序，对于研究强化学习和自适应动态规划的朋友很有参考价值(Q learning MATLAB. For the study of reinforcement learning and adaptive dynamic programming )

一系列展示图像压缩技术的源代码。包括有：使用块截断编码的图像压缩(Block Truncation)、基于高斯金字塔变换的图像压缩(Gaussian Pyramids)、基于离散余弦变换对图像压缩(Discrete Cosine Transform)、基于奇异值分解(SVD)的图像压缩(Singular Value Decomposition)。给出的代码还可以用于2D图像噪声消除。(Image Compression A collection of simple routines for image compression using different techniques. BTCODE: Image compression Using Block Truncation Coding. PYRAMID: Image compression based on Gaussian Pyramids. DCTCOMPR: Image compression based on Discrete Cosine Transform. IMCOMPR: Image compression based on Singular Value Decomposition. The given codes can be also used in 2D noise suppression. Notes: The function "conv2fft" performs a 2D FFT-based convolution. Type "help conv2fft" on Matlab command window for more informations.)

说明：  采用白噪声作为输入信号，用递推最小二乘法来进行系统辨识(Using white noise as input signal, using recursive least squares method for system identification)

此源码包含了QPSK调制所需的一系列子程序，方便大家学习参考(This source contains a series of subroutines required for QPSK modulation to facilitate them to learn from reference)

线性系统的状态空间分析与综合，程序均有详细注释(The state space of the linear system analysis and synthesis procedures have detailed notes)

基于小波的脑部MR与伪彩色PET医学图像融合，低频采用加权平均，高频采用3*3区域标准差的融合规则，效果很好，内含一组脑部MR与PET医学图像(Simulation Wavelet-based for brain PET and MR image fusion, low frequency using the weighted average and high frequency using region-based standard deviation fusion rule, the effect is very good, including a set of MR and PET images)

功能: 精确线搜索之抛物线法 输入: phi 是目标函数, a和b是搜索区间的端点(Features: accurate line search of the parabola method input: phi is the objective function, a and b are the endpoints of the search range)

matlab simulationfive least source

包含多种噪声谱估计算法用于语音增强，采用的是标准谱减法，主要用于对各种噪声谱估计算法对比，详见readme.txt(Ppls spectral subtraction noise spectrum estimation, mainly used to compare the performance of different noise spectrum estimation algorithms, including Martin s minimum tracking algorithm, Minimum controlled recursive average algorithm (Cohen), variant of Minimum controlled recursive average algorithm, improved Minimum controlled recursive average algorithm (Cohen ), continuous spectral minimum tracking (Doblinger), weighted spectral average (Hirsch & Ehrilcher), connected frequency regions (Sorensen & Andersen). See readme.txt inside)