LSSVM.rar_ls-svm_LSSVM matlab_行业应用下载-pudn.com

<html xmlns="http://www.w3.org/1999/xhtml"> <head> <meta charset="utf-8"> <meta name="generator" content="pdf2htmlEX"> <meta http-equiv="X-UA-Compatible" content="IE=edge,chrome=1"> <link rel="stylesheet" href="https://static.pudn.com/base/css/base.min.css"> <link rel="stylesheet" href="https://static.pudn.com/base/css/fancy.min.css"> <link rel="stylesheet" href="https://static.pudn.com/prod/directory_preview_static/622b588d15da9b288bcf10c0/raw.css"> <script src="https://static.pudn.com/base/js/compatibility.min.js"></script> <script src="https://static.pudn.com/base/js/pdf2htmlEX.min.js"></script> <script> try{ pdf2htmlEX.defaultViewer = new pdf2htmlEX.Viewer({}); }catch(e){} </script> <title></title> </head> <body> <div id="sidebar" style="display: none"> <div id="outline"> </div> </div> <div id="pf1" class="pf w0 h0" data-page-no="1"><div class="pc pc1 w0 h0"><img class="bi x0 y0 w1 h1" alt="" src="https://static.pudn.com/prod/directory_preview_static/622b588d15da9b288bcf10c0/bg1.jpg"><div class="t m0 x1 h2 y1 ff1 fs0 fc0 sc0 ls0 ws0">LS-SVMlab Toolbox User’s Guide</div><div class="t m0 x2 h3 y2 ff2 fs1 fc0 sc0 ls0 ws0">version 1.5</div><div class="t m0 x3 h4 y3 ff1 fs2 fc0 sc0 ls0 ws0">K. Pelckmans, J.A.K. Suykens, T. Van Gestel, J. De Brabanter,</div><div class="t m0 x4 h4 y4 ff1 fs2 fc0 sc0 ls0 ws0">L. Lukas, B. Hamers, B. De Moor, J. Vandewalle</div><div class="t m0 x5 h5 y5 ff3 fs2 fc0 sc0 ls0 ws0">Katholieke Universiteit Leuven</div><div class="t m0 x6 h5 y6 ff3 fs2 fc0 sc0 ls0 ws0">Department of Electrical Engineering, ESAT-SCD-SISTA</div><div class="t m0 x7 h5 y7 ff3 fs2 fc0 sc0 ls0 ws0">Kasteelpark Arenberg 10, B-3001 Leuven-Heverlee, Belgium</div><div class="t m0 x8 h4 y8 ff4 fs2 fc0 sc0 ls0 ws0">{ kristiaan.pelckmans, johan.suykens }@esat.kuleuven.ac.be</div><div class="t m0 x9 h4 y9 ff1 fs2 fc0 sc0 ls0 ws0">http://www.esat.kuleuven.ac.be/sista/lssvmlab/</div><div class="t m0 xa h5 ya ff3 fs2 fc0 sc0 ls0 ws0">ESAT-SCD-SISTA Technical Report 02-145</div><div class="t m0 x2 h5 yb ff3 fs2 fc0 sc0 ls0 ws0">February 2003</div></div><div class="pi" data-data='{"ctm":[1.568627,0.000000,0.000000,1.568627,0.000000,0.000000]}'></div></div> </body> </html>

<div id="pf2" class="pf w0 h0" data-page-no="2"><div class="pc pc2 w0 h0"><img class="bi x0 y0 w1 h1" alt="" src="https://static.pudn.com/prod/directory_preview_static/622b588d15da9b288bcf10c0/bg2.jpg"><div class="t m0 xb h6 yc ff5 fs3 fc0 sc0 ls0 ws0">Acknowledgements</div><div class="t m0 xc h6 yd ff6 fs3 fc0 sc0 ls0 ws0">Research supported by Research Council K.U.Leuven: GOA-Meﬁsto 666, IDO (IOTA</div><div class="t m0 xc h6 ye ff6 fs3 fc0 sc0 ls0 ws0">oncology, genetic networks), several PhD/postdoc & fellow grants; Flemish Govern-</div><div class="t m0 xc h6 yf ff7 fs3 fc0 sc0 ls0 ws0">ment: FWO: PhD/postdoc grants, G.0407.02 (support vector machines), projects</div><div class="t m0 xc h6 y10 ff6 fs3 fc0 sc0 ls0 ws0">G.0115.01 (microarrays/oncology), G.0240.99 (multilinear algebra), G.0080.01 (col-</div><div class="t m0 xc h6 y11 ff6 fs3 fc0 sc0 ls0 ws0">lective intelligence), G.0413.03 (inference in bioi), G.0388.03 (microarrays for clinical</div><div class="t m0 xc h6 y12 ff6 fs3 fc0 sc0 ls0 ws0">use), G.0229.03 (ontologies in bioi), G.0197.02 (power islands), G.0141.03 (identiﬁ-</div><div class="t m0 xc h6 y13 ff6 fs3 fc0 sc0 ls0 ws0">cation and cryptography), G.0491.03 (control for intensive care glycemia), G.0120.03</div><div class="t m0 xc h6 y14 ff6 fs3 fc0 sc0 ls0 ws0">(QIT), research communities (ICCoS, ANMMM); AWI: Bil. Int. Collaboration Hun-</div><div class="t m0 xc h6 y15 ff6 fs3 fc0 sc0 ls0 ws0">gary, Poland, South Africa; IWT: PhD G rants, STWW-Genprom (gene promotor</div><div class="t m0 xc h6 y16 ff6 fs3 fc0 sc0 ls0 ws0">prediction), GBOU-McKnow (knowledge management algorithms), GBOU-SQUAD</div><div class="t m0 xc h6 y17 ff6 fs3 fc0 sc0 ls0 ws0">(quorum sensing), GBOU-ANA (biosensors); Soft4s (softsensors) Belgian Federal Gov-</div><div class="t m0 xc h6 y18 ff7 fs3 fc0 sc0 ls0 ws0">ernment: DWTC (IUAP IV-02 (1996-2001) and IUAP V-22 (2002-2006)); PODO-II</div><div class="t m0 xc h6 y19 ff6 fs3 fc0 sc0 ls0 ws0">(CP/40: TMS and sustainibility); EU: CAGE; ERNSI; Eureka 2063-IMPACT; Eureka</div><div class="t m0 xc h6 y1a ff6 fs3 fc0 sc0 ls0 ws0">2419-FliTE; Contract Research/agreements: Data4s, Electrabel, Elia, LMS, IPCOS,</div><div class="t m0 xc h6 y1b ff6 fs3 fc0 sc0 ls0 ws0">VIB; JS is a professor at K.U.Leuven Belgium and a postdoctoral researcher with FWO</div><div class="t m0 xc h6 y1c ff6 fs3 fc0 sc0 ls0 ws0">Flanders. TVG is postdoctoral researcher with FWO Flanders. BDM and JWDW are</div><div class="t m0 xc h6 y1d ff6 fs3 fc0 sc0 ls0 ws0">full professors at K.U.Leuven Belgium.</div><div class="t m0 xd h6 y1e ff6 fs3 fc0 sc0 ls0 ws0">1</div></div><div class="pi" data-data='{"ctm":[1.568627,0.000000,0.000000,1.568627,0.000000,0.000000]}'></div></div>

<div id="pf3" class="pf w0 h0" data-page-no="3"><div class="pc pc3 w0 h0"><img class="bi x0 y0 w1 h1" alt="" src="https://static.pudn.com/prod/directory_preview_static/622b588d15da9b288bcf10c0/bg3.jpg"><div class="t m0 xe h7 y1f ff1 fs4 fc0 sc0 ls0 ws0">Contents</div><div class="t m0 xe h6 y20 ff5 fs3 fc0 sc0 ls0 ws0">1 Introduction 4</div><div class="t m0 xe h6 y21 ff5 fs3 fc0 sc0 ls0 ws0">2 A birds eye view on LS-SVMlab 5</div><div class="t m0 xf h6 y22 ff6 fs3 fc0 sc0 ls0 ws0">2.1 Classiﬁcation and Regression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5</div><div class="t m0 x1 h6 y23 ff6 fs3 fc0 sc0 ls0 ws0">2.1.1 Classiﬁcation Extensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6</div><div class="t m0 x1 h6 y24 ff6 fs3 fc0 sc0 ls0 ws0">2.1.2 Tuning, Sparseness, Robustness . . . . . . . . . . . . . . . . . . . . . . . . . 6</div><div class="t m0 x1 h6 y25 ff6 fs3 fc0 sc0 ls0 ws0">2.1.3 Bayesian Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8</div><div class="t m0 xf h6 y26 ff6 fs3 fc0 sc0 ls0 ws0">2.2 NARX Models and Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8</div><div class="t m0 xf h6 y27 ff6 fs3 fc0 sc0 ls0 ws0">2.3 Unsupervised Learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8</div><div class="t m0 xf h6 y28 ff6 fs3 fc0 sc0 ls0 ws0">2.4 Solving Large Scale Problems with Fixed Size LS-SVM . . . . . . . . . . . . . . . . 9</div><div class="t m0 xe h6 y29 ff5 fs3 fc0 sc0 ls0 ws0">3 LS-SVMlab toolbox examples 10</div><div class="t m0 xf h6 y2a ff6 fs3 fc0 sc0 ls0 ws0">3.1 Classiﬁcation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10</div><div class="t m0 x1 h6 y2b ff6 fs3 fc0 sc0 ls0 ws0">3.1.1 Hello world... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10</div><div class="t m0 x1 h6 y2c ff6 fs3 fc0 sc0 ls0 ws0">3.1.2 The Ripley data set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12</div><div class="t m0 x1 h6 y2d ff6 fs3 fc0 sc0 ls0 ws0">3.1.3 Bayesian Inference for Classiﬁcation . . . . . . . . . . . . . . . . . . . . . . 14</div><div class="t m0 x1 h6 y2e ff6 fs3 fc0 sc0 ls0 ws0">3.1.4 Multi-class co ding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16</div><div class="t m0 xf h6 y2f ff6 fs3 fc0 sc0 ls0 ws0">3.2 Regression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17</div><div class="t m0 x1 h6 y30 ff6 fs3 fc0 sc0 ls0 ws0">3.2.1 A Simple Sinc Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17</div><div class="t m0 x1 h6 y31 ff6 fs3 fc0 sc0 ls0 ws0">3.2.2 Bayesian Inference for Regression . . . . . . . . . . . . . . . . . . . . . . . . 19</div><div class="t m0 x1 h6 y32 ff6 fs3 fc0 sc0 ls0 ws0">3.2.3 Using the object oriented model interface . . . . . . . . . . . . . . . . . . . 20</div><div class="t m0 x1 h6 y33 ff6 fs3 fc0 sc0 ls0 ws0">3.2.4 Robust Regression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22</div><div class="t m0 x1 h6 y34 ff6 fs3 fc0 sc0 ls0 ws0">3.2.5 Multiple Output Regression . . . . . . . . . . . . . . . . . . . . . . . . . . . 23</div><div class="t m0 x1 h6 y35 ff6 fs3 fc0 sc0 ls0 ws0">3.2.6 A Time-Series Example: Santa Fe Laser Data Prediction . . . . . . . . . . 24</div><div class="t m0 x1 h6 y36 ff6 fs3 fc0 sc0 ls0 ws0">3.2.7 Fixed size LS-SVM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25</div><div class="t m0 xf h6 y37 ff6 fs3 fc0 sc0 ls0 ws0">3.3 Unsupervised Learning using kernel based Principal Component Analysis . . . . . 28</div><div class="t m0 xe h6 y38 ff5 fs3 fc0 sc0 ls0 ws0">A MATLAB functions 29</div><div class="t m0 xf h6 y39 ff6 fs3 fc0 sc0 ls0 ws0">A.1 General Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29</div><div class="t m0 xf h6 y3a ff6 fs3 fc0 sc0 ls0 ws0">A.2 Index of Function Calls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30</div><div class="t m0 x1 h6 y3b ff6 fs3 fc0 sc0 ls0 ws0">A.2.1 Training and Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30</div><div class="t m0 x1 h6 y3c ff6 fs3 fc0 sc0 ls0 ws0">A.2.2 Object Oriented Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31</div><div class="t m0 x1 h6 y3d ff6 fs3 fc0 sc0 ls0 ws0">A.2.3 Training and Simulating Functions . . . . . . . . . . . . . . . . . . . . . . . 32</div><div class="t m0 x1 h6 y3e ff6 fs3 fc0 sc0 ls0 ws0">A.2.4 Kernel Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33</div><div class="t m0 x1 h6 y3f ff6 fs3 fc0 sc0 ls0 ws0">A.2.5 Tuning, Sparseness and Robustness . . . . . . . . . . . . . . . . . . . . . . . 34</div><div class="t m0 x1 h6 y40 ff6 fs3 fc0 sc0 ls0 ws0">A.2.6 Classiﬁcation Extensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35</div><div class="t m0 x1 h6 y41 ff6 fs3 fc0 sc0 ls0 ws0">A.2.7 Bayesian Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36</div><div class="t m0 x1 h6 y42 ff6 fs3 fc0 sc0 ls0 ws0">A.2.8 NARX mo dels and Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . 37</div><div class="t m0 x1 h6 y43 ff6 fs3 fc0 sc0 ls0 ws0">A.2.9 Unsupervised learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38</div><div class="t m0 x1 h6 y44 ff6 fs3 fc0 sc0 ls0 ws0">A.2.10 Fixed Size LS-SVM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39</div><div class="t m0 x1 h6 y45 ff6 fs3 fc0 sc0 ls0 ws0">A.2.11 Demos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40</div><div class="t m0 xf h6 y46 ff6 fs3 fc0 sc0 ls0 ws0">A.3 Alphabetical List of Function Calls . . . . . . . . . . . . . . . . . . . . . . . . . . . 41</div><div class="t m0 xd h6 y47 ff6 fs3 fc0 sc0 ls0 ws0">2</div></div><div class="pi" data-data='{"ctm":[1.568627,0.000000,0.000000,1.568627,0.000000,0.000000]}'></div></div>

<div id="pf4" class="pf w0 h0" data-page-no="4"><div class="pc pc4 w0 h0"><img class="bi x0 y0 w1 h1" alt="" src="https://static.pudn.com/prod/directory_preview_static/622b588d15da9b288bcf10c0/bg4.jpg"><div class="t m0 x1 h6 yc ff6 fs3 fc0 sc0 ls0 ws0">A.3.1 AFE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41</div><div class="t m0 x1 h6 y48 ff6 fs3 fc0 sc0 ls0 ws0">A.3.2 bay</div><div class="t m0 x10 h6 y49 ff9 fs3 fc0 sc0 ls0 ws0">errorbar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42</div><div class="t m0 x1 h6 y4a ff6 fs3 fc0 sc0 ls0 ws0">A.3.3 bay initlssvm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44</div><div class="t m0 x1 h6 y4b ff6 fs3 fc0 sc0 ls0 ws0">A.3.4 bay lssvm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45</div><div class="t m0 x1 h6 y4c ff6 fs3 fc0 sc0 ls0 ws0">A.3.5 bay lssvmARD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47</div><div class="t m0 x1 h6 y4d ff6 fs3 fc0 sc0 ls0 ws0">A.3.6 bay modoutClass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49</div><div class="t m0 x1 h6 y4e ff6 fs3 fc0 sc0 ls0 ws0">A.3.7 bay optimize . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51</div><div class="t m0 x1 h6 y4f ff6 fs3 fc0 sc0 ls0 ws0">A.3.8 bay rr . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53</div><div class="t m0 x1 h6 y50 ff6 fs3 fc0 sc0 ls0 ws0">A.3.9 code, codelssvm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55</div><div class="t m0 x1 h6 y51 ff6 fs3 fc0 sc0 ls0 ws0">A.3.10 crossvalidate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58</div><div class="t m0 x1 h6 y52 ff6 fs3 fc0 sc0 ls0 ws0">A.3.11 deltablssvm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61</div><div class="t m0 x1 h6 y53 ff6 fs3 fc0 sc0 ls0 ws0">A.3.12 denoise kpca . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62</div><div class="t m0 x1 h6 y54 ff6 fs3 fc0 sc0 ls0 ws0">A.3.13 eign . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64</div><div class="t m0 x1 h6 y55 ff6 fs3 fc0 sc0 ls0 ws0">A.3.14 initlssvm, changelssvm . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65</div><div class="t m0 x1 h6 y56 ff6 fs3 fc0 sc0 ls0 ws0">A.3.15 kentropy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68</div><div class="t m0 x1 h6 y57 ff6 fs3 fc0 sc0 ls0 ws0">A.3.16 kernel</div><div class="t m0 x11 h6 y58 ff9 fs3 fc0 sc0 ls0 ws0">matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69</div><div class="t m0 x1 h6 y59 ff6 fs3 fc0 sc0 ls0 ws0">A.3.17 kpca . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70</div><div class="t m0 x1 h6 y5a ff6 fs3 fc0 sc0 ls0 ws0">A.3.18 latentlssvm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71</div><div class="t m0 x1 h6 y5b ff6 fs3 fc0 sc0 ls0 ws0">A.3.19 leaveoneout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73</div><div class="t m0 x1 h6 y5c ff6 fs3 fc0 sc0 ls0 ws0">A.3.20 leaveoneout</div><div class="t m0 x12 h6 y5d ff9 fs3 fc0 sc0 ls0 ws0">lssvm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75</div><div class="t m0 x1 h6 y5e ff6 fs3 fc0 sc0 ls0 ws0">A.3.21 lin kernel, MLP kernel, poly kernel, RBF kernel . . . . . . . . . . . . 77</div><div class="t m0 x1 h6 y5f ff6 fs3 fc0 sc0 ls0 ws0">A.3.22 linf, mae, medae, misclass, mse, trimmedmse . . . . . . . . . . . . . 78</div><div class="t m0 x1 h6 y60 ff6 fs3 fc0 sc0 ls0 ws0">A.3.23 plotlssvm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80</div><div class="t m0 x1 h6 y61 ff6 fs3 fc0 sc0 ls0 ws0">A.3.24 predict . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82</div><div class="t m0 x1 h6 y62 ff6 fs3 fc0 sc0 ls0 ws0">A.3.25 prelssvm, postlssvm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84</div><div class="t m0 x1 h6 y63 ff6 fs3 fc0 sc0 ls0 ws0">A.3.26 rcrossvalidate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85</div><div class="t m0 x1 h6 y64 ff6 fs3 fc0 sc0 ls0 ws0">A.3.27 ridgeregress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87</div><div class="t m0 x1 h6 y65 ff6 fs3 fc0 sc0 ls0 ws0">A.3.28 robustlssvm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88</div><div class="t m0 x1 h6 y66 ff6 fs3 fc0 sc0 ls0 ws0">A.3.29 roc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89</div><div class="t m0 x1 h6 y67 ff6 fs3 fc0 sc0 ls0 ws0">A.3.30 simlssvm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91</div><div class="t m0 x1 h6 y68 ff6 fs3 fc0 sc0 ls0 ws0">A.3.31 sparselssvm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93</div><div class="t m0 x1 h6 y69 ff6 fs3 fc0 sc0 ls0 ws0">A.3.32 trainlssvm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94</div><div class="t m0 x1 h6 y6a ff6 fs3 fc0 sc0 ls0 ws0">A.3.33 tunelssvm, linesearch & gridsearch . . . . . . . . . . . . . . . . . . . . 96</div><div class="t m0 x1 h6 y6b ff6 fs3 fc0 sc0 ls0 ws0">A.3.34 validate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100</div><div class="t m0 x1 h6 y6c ff6 fs3 fc0 sc0 ls0 ws0">A.3.35 windowize & windowizeNARX . . . . . . . . . . . . . . . . . . . . . . . . . . 102</div><div class="t m0 xd h6 y6d ff6 fs3 fc0 sc0 ls0 ws0">3</div></div><div class="pi" data-data='{"ctm":[1.568627,0.000000,0.000000,1.568627,0.000000,0.000000]}'></div></div>

<div id="pf5" class="pf w0 h0" data-page-no="5"><div class="pc pc5 w0 h0"><img class="bi x0 y0 w1 h1" alt="" src="https://static.pudn.com/prod/directory_preview_static/622b588d15da9b288bcf10c0/bg5.jpg"><div class="t m0 xe h2 y6e ff1 fs0 fc0 sc0 ls0 ws0">Chapter 1</div><div class="t m0 xe h7 y6f ff1 fs4 fc0 sc0 ls0 ws0">Introduction</div><div class="t m0 xe h6 y70 ff6 fs3 fc0 sc0 ls0 ws0">Support Vector Machines (SVM) is a powerful methodology for solving problems in nonlinear</div><div class="t m0 xe h6 y71 ff6 fs3 fc0 sc0 ls0 ws0">classiﬁcation, function estimation and density estimation which has also led to many other recent</div><div class="t m0 xe h6 y72 ff6 fs3 fc0 sc0 ls0 ws0">developments in kernel based learning methods in general [3, 16, 17, 34, 33]. SVMs have been in-</div><div class="t m0 xe h6 y73 ff6 fs3 fc0 sc0 ls0 ws0">troduced within the context of statistical learning theory and structural risk minimization. In the</div><div class="t m0 xe h6 y74 ff6 fs3 fc0 sc0 ls0 ws0">methods one solves convex optimization problems, typically quadratic programs. Least Squares</div><div class="t m0 xe h6 y75 ff6 fs3 fc0 sc0 ls0 ws0">Support Vector Machines (LS-SVM) are reformulations to standard SVMs [21, 28] which lead</div><div class="t m0 xe h6 y76 ff6 fs3 fc0 sc0 ls0 ws0">to solving linear KKT systems. LS-SVMs are closely related to regularization networks [5] and</div><div class="t m0 xe h6 y77 ff6 fs3 fc0 sc0 ls0 ws0">Gaussian processes [37] but additionally emphasize and exploit primal-dual interpretations. Links</div><div class="t m0 xe h6 y78 ff6 fs3 fc0 sc0 ls0 ws0">between kernel versions of classical pattern recognition algorithms such as kernel Fisher discrim-</div><div class="t m0 xe h6 y79 ff6 fs3 fc0 sc0 ls0 ws0">inant analysis and extensions to unsupervised learning, recurrent networks and control [22] are</div><div class="t m0 xe h6 y7a ff6 fs3 fc0 sc0 ls0 ws0">available. Robustness, sparseness and weightings [23] can be imposed to LS-SVMs where needed</div><div class="t m0 xe h6 y7b ff6 fs3 fc0 sc0 ls0 ws0">and a Bayesian framework with three levels of inference has been developed [29, 32]. LS-SVM</div><div class="t m0 xe h6 y7c ff6 fs3 fc0 sc0 ls0 ws0">alike primal-dual formulations are given to kernel PCA [24], kernel CCA and kernel PLS [25]. For</div><div class="t m0 xe h6 y7d ff6 fs3 fc0 sc0 ls0 ws0">ultra large scale problems and on-line learning a method of Fixed Size LS-SVM is proposed, which</div><div class="t m0 xe h6 y7e ff6 fs3 fc0 sc0 ls0 ws0">is related to a Nystr¨om sampling [6, 35] with active selection of support vectors and estimation in</div><div class="t m0 xe h6 y7f ff6 fs3 fc0 sc0 ls0 ws0">the primal space.</div><div class="t m0 xf h6 y80 ff6 fs3 fc0 sc0 ls0 ws0">The present LS-SVMlab toolbox User’s G uide contains Matlab/C implementations for a num-</div><div class="t m0 xe h6 y81 ff6 fs3 fc0 sc0 ls0 ws0">ber of LS-SVM algorithms related to classiﬁcation, regression, time-series prediction and unsuper-</div><div class="t m0 xe h6 y82 ff6 fs3 fc0 sc0 ls0 ws0">vised learning. References to commands in the toolbox are written in typewriter font.</div><div class="t m0 xf h6 y83 ff6 fs3 fc0 sc0 ls0 ws0">A main reference and overview on least squares support vector machines is</div><div class="t m0 xc h6 y84 ff6 fs3 fc0 sc0 ls0 ws0">J.A.K. Suykens, T. Van Gestel, J. De Brabanter, B. De Moor, J. Vandewalle,</div><div class="t m0 xc h6 y85 ffa fs3 fc0 sc0 ls0 ws0">Least Squares Support Vector Machines,</div><div class="t m0 xc h6 y86 ff6 fs3 fc0 sc0 ls0 ws0">World Scientiﬁc, Singapore, 2002 (ISBN 981-238-151-1).</div><div class="t m0 xf h6 y87 ff6 fs3 fc0 sc0 ls0 ws0">The LS-SVMlab homepage is</div><div class="t m0 x13 h8 y88 ff9 fs3 fc0 sc0 ls0 ws0">http://www.esat.kuleuven.ac.be/sista/lssvmlab/</div><div class="t m0 xf h6 y89 ff6 fs3 fc0 sc0 ls0 ws0">The LS-SVMlab to olbox is made available under the GNU general license p olicy:</div><div class="t m0 xc h6 y8a ff7 fs3 fc0 sc0 ls0 ws0">Copyright (C) 2002 KULeuven-ESAT-SCD</div><div class="t m0 xc h6 y8b ff7 fs3 fc0 sc0 ls0 ws0">This program is free software; you can redistribute it and/or modify it under the terms</div><div class="t m0 xc h6 y8c ff7 fs3 fc0 sc0 ls0 ws0">of the GNU General Public License as published by the Free Software Foundation;</div><div class="t m0 xc h6 y8d ff7 fs3 fc0 sc0 ls0 ws0">either version 2 of the License, or (at your option) any later version.</div><div class="t m0 xc h6 y8e ff7 fs3 fc0 sc0 ls0 ws0">This program is distributed in the hope that it will be useful, but WITHOUT ANY</div><div class="t m0 xc h6 y8f ff7 fs3 fc0 sc0 ls0 ws0">WARRANTY; without even the implied warranty of MERCHANTABILITY or FIT-</div><div class="t m0 xc h6 y90 ff7 fs3 fc0 sc0 ls0 ws0">NESS FOR A PARTICULAR PURPOSE. See the website of LS-SVMlab or the GNU</div><div class="t m0 xc h6 y91 ff7 fs3 fc0 sc0 ls0 ws0">General Public License for a copy of the GNU General Public License speciﬁcations.</div><div class="t m0 xd h6 y1e ff6 fs3 fc0 sc0 ls0 ws0">4</div></div><div class="pi" data-data='{"ctm":[1.568627,0.000000,0.000000,1.568627,0.000000,0.000000]}'></div></div>