function[PSNR,y_est]=BM3D(y,z,sigma,profile,print_to_screen,N2,N2_wiener,Ns,Ns_wiener,tau_match)%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%BM3DisanalgorithmforattenuationofadditivewhiteGaussiannoisefrom%grayscaleimages.Thisalgorithmreproducestheresultsfromthearticle:%%[1]K.Dabov,A.Foi,V.Katkovnik,andK.Egiazarian,"ImageDenoising%bySparse3DTransform-DomainCollaborativeFiltering,"%IEEETransactionsonImageProcessing,vol.16,no.8,August,2007.%preprintathttp://www.cs.tut.fi/~foi/GCF-BM3D.%%%FUNCTIONINTERFACE:%%[PSNR,y_est]=BM3D(y,z,sigma,profile,print_to_screen)%%!Thefunctioncanworkwithoutanyoftheinputarguments,%inwhichcase,theinternaldefaultonesareused!%%BASICUSAGEEXAMPLES:%%Case1)Usingthedefaultparameters(i.e.,imagename,sigma,etc.)%%[PSNR,y_est]=BM3D;%%Case2)Usinganexternalnoisyimage:

TianyiZhou，DachengTao等人提出的GoDec模型，适用于低秩分解。

Thistextbookintroducessparseandredundantrepresentationswithafocusonapplicationsinsignalandimageprocessing.Thetheoreticalandnumericalfoundationsaretackledbeforetheapplicationsarediscussed.Mathematicalmodelingforsignalsourcesisdiscussedalongwithhowtousethepropermodelfortaskssuchasdenoising,restoration,separation,interpolationandextrapolation,compression,sampling,analysisandsynthesis,detection,recognition,andmore.Thepresentationiselegantandengaging.SparseandRedundantRepresentationsisintendedforgraduatestudentsinappliedmathematicsandelectricalengineering,aswellasappliedmathematicians,engineers,andresearcherswhoareactiveinthefieldsofsignalandimageprocessing.*Introducestheoreticalandnumericalfoundationsbeforetacklingapplications*Discusseshowtousethepropermodelforvarioussituations*Introducessparseandredundantrepresentations*FocusesonapplicationsinsignalandimageprocessingThefieldofsparseandredundantrepresentationmodelinghasgonethroughamajorrevolutioninthepasttwodecades.Thisstartedwithaseriesofalgorithmsforapproximatingthesparsestsolutionsoflinearsystemsofequations,latertobefollowedbysurprisingtheoreticalresultsthatguaranteethesealgorithms’performance.Withthesecontributionsinplace,majorbarriersinmakingthismodelpracticalandapplicablewereremoved,andsparsityandredundancybecamecentral,leadingtostate-of-the-artresultsinvariousdisciplines.Oneofthemainbeneficiariesofthisprogressisthefieldofimageprocessing,wherethismodelhasbeenshowntoleadtounprecedentedperformanceinvariousapplications.Thisbookprovidesacomprehensiveviewofthetopicofsparseandredundantrepresentationmodeling,anditsuseinsignalandimageprocessing.Itoffersasystematicandorderedexposuretothetheoreticalfoundationsofthisdatamodel,thenumericalaspec

稀疏直接求解法的好教材,LU分解法，对称和非对称求解法

nmanydataanalysistasks,oneisoftenconfrontedwithveryhighdimensionaldata.Featureselectiontechniquesaredesignedtofindtherelevantfeaturesubsetoftheoriginalfeatureswhichcanfacilitateclustering,classificationandretrieval.Thefeatureselectionproblemisessentiallyacombinatorialoptimizationproblemwhichiscomputationallyexpensive.Traditionalfeatureselectionmethodsaddressthisissuebyselectingthetoprankedfeaturesbasedoncertainscorescomputedindependentlyforeachfeature.Theseapproachesneglectthepossiblecorrelationbetweendifferentfeaturesandthuscannotproduceanoptimalfeaturesubset.InspiredfromtherecentdevelopmentsonmanifoldlearningandL1-regularizedmodelsforsubsetselection,weproposehereanewapproach,called{\emMulti-Cluster/ClassFeatureSelection}(MCFS),forfeatureselection.Specifically,weselectthosefeaturessuchthatthemulti-cluster/classstructureofthedatacanbebestpreserved.Thecorrespondingoptimizationproblemcanbeefficientlysolvedsinceitonlyinvolvesasparseeigen-problemandaL1-regularizedleastsquaresproblem.ItisimportanttonotethatMCFScanbeappliedinsuperised,unsupervisedandsemi-supervisedcases.Ifyoufindthesealgoirthmsuseful,weappreciateitverymuchifyoucanciteourfollowingworks:PapersDengCai,ChiyuanZhang,XiaofeiHe,"UnsupervisedFeatureSelectionforMulti-clusterData",16thACMSIGKDDConferenceonKnowledgeDiscoveryandDataMining(KDD'10),July2010.BibtexsourceXiaofeiHe,DengCai,andParthaNiyogi,"LaplacianScoreforFeatureSelection",AdvancesinNeuralInformationProcessingSystems18(NIPS'05),Vancouver,Canada,2005Bibtexsource

SuiteSparse是世界上最优秀的系数矩阵处理工程之一。
但是SuiteSparse提供的官方代码仅包含在matlab、linux环境下编译的生成文件，不能生成在windows操作系统下VS环境下的C++库函数。
本文件包括一个库函数cs.cpp和一个头文件cs.h，其中的代码是移植自SuiteSparse官方代码中的Csparse原始代码，功能包括除了复数矩阵以外的所有功能，已成功在vs2010的c++环境下执行过，在毕业设计中用于求解超大型稀疏矩阵的线性方程组（也就是大型稀疏矩阵的除法）。
以下是SuiteSparse的介绍。
SuiteSparse是一组C、Fortran和MATLAB函数集，用来生成空间稀疏矩阵数据。
在SuiteSparse中几何多种稀疏矩阵的处理方法，包括矩阵的LU分解，QR分解，Cholesky分解，提供了解非线性方程组、实现最小二乘法等多种函数代码。

Thisbookcoversalgorithmicandhardwareimplementationtechniquestoenableembeddeddeeplearning.Theauthorsdescribesynergeticdesignapproachesontheapplication-,algorithmic-,computerarchitecture-,andcircuit-levelthatwillhelpinachievingthegoalofreducingthecomputationalcostofdeeplearningalgorithms.Theimpactofthesetechniquesisdisplayedinfoursiliconprototypesforembeddeddeeplearning.Givesawideoverviewofaseriesofeffectivesolutionsforenergy-efficientneuralnetworksonbatteryconstrainedwearabledevices;Discussestheoptimizationofneuralnetworksforembeddeddeploymentonalllevelsofthedesignhierarchy–applications,algorithms,hardwarearchitectures,andcircuits–supportedbyrealsiliconprototypes;ElaboratesonhowtodesignefficientConvolutionalNeuralNetworkprocessors,exploitingparallelismanddata-reuse,sparseoperations,andlow-precisioncomputations;Supportstheintroducedtheoryanddesignconceptsbyfourrealsiliconprototypes.Thephysicalrealization’simplementationandachievedperformancesarediscussedelaboratelytoillustratedandhighlighttheintroducedcross-layerdesignconcepts.

压缩感知offgrid代码论文见2013SparseFrequencydiverseMIMOradarimagingforOff-GridtargetbasedonadaptiveiterativeMAP

该程序是以色列科学家michael关于稀疏表示字典训练的一种算法仿真，并附有应用该字典进行图象去噪的实例.应用该程序可以按照readme中的提示一步一步完成，先安装sparse-codingoflargesetsofsignals文件夹中的内容,再安装ksvd程序就可以运行了.

SuiteSparse的vs2017库

在日常工作中，钉钉打卡成了我生活中不可或缺的一部分。然而，有时候这个看似简单的任务却给我带来了不少烦恼。 每天早晚，我总是得牢记打开钉钉应用，点击"工作台"，再找到"考勤打卡"进行签到。有时候因为工作忙碌，会忘记打卡，导致考勤异常，影响当月的工作评价。而且，由于我使用的是苹果手机，有时候系统更新后，钉钉的某些功能会出现异常，使得打卡变得更加麻烦。 另外，我的家人使用的是安卓手机，他们也经常抱怨钉钉打卡的繁琐。尤其是对于那些不太熟悉手机操作的长辈来说，每次打卡都是一次挑战。他们总是担心自己会操作失误，导致打卡失败。 为了解决这些烦恼，我开始思考是否可以通过编写一个全自动化脚本来实现钉钉打卡。经过一段时间的摸索和学习，我终于成功编写出了一个适用于苹果和安卓系统的钉钉打卡脚本。