ELF<4 N4 (444JJJpJ((( Qtd/lib/ld-linux.so.2GNU#  "!    YV !#J (!kr]$6' }1)3, z1 R0libm.so.6sqrttanhpowexp_Jv_RegisterClasses__gmon_start__libc.so.6putcharstrcpy__strtod_internalstdout__strtol_internalfscanffgetsmemcpyperrorputsfeofmallocfflushfprintfstdin_IO_getcsscanffcloseisnan__ctype_b_locfwriteexitfopen_IO_stdin_used__libc_start_mainclockfreeGLIBC_2.0GLIBC_2.3GLIBC_2.1 ii .@ii 8ii Bii .!    $ ( ,048<@DHLPTX\`dh l"Ud5%%h%h%h%h% h %h(%h0%h8p%h@`% hHP%$hP@%(hX0%,h` %0hh%4hp%8hx%<h%@h%Dh%Hh%Lh%Ph%Th%Xhp%\h`%`hP%dh@%hh0%lh 1^PTRhhLQVh|US[ÇQRtX[ÐU=tvxҡxuÉUtt hЃvÐUh0X5 ÉU h8hAh^$@$G$x$$R$$ $cy$@m$a$؛U$I$P=$1$%$$H $$ȝ$$P$$$$$T$$$ }$,q$te$Y$uM$A$D5$p)$$$ $`$$Ԣ$$L$$$$8$$$u$4i$|]$Q$E$D9$-$ئ!$ $h $$$L$$$$($l$$$0Z5$m$<a$U$ԪI$=$D1$%$$ $,$|$Ĭ$$P$ÍvUWVSEsvm_@modef@l] h߾tf] ƃ0Mu11?S{C C$C(CPCTC, C0CHCLC4(KCC 1ɋu ?N^ h㈋u $>ҺMbP?(F8Vjj jM F4UE F4E$ P1F} 4xPFU 4u PhL/Ph[^U r0hد Ph؃ hh$؃ h؃ hܰ؃ h hH$$ h0xUWVShHDž0.VWhhhhhu uk$DP@Q؋U0JuUEEu%U$B u M $]}0_EE]E(<UEuE@؋]0sM0JAuI44ыTTًUMvE]E(<p؋}$O$tMuu0J;V=hU0e[^H_ÐE]M0AE@t"؋u0FM0] q (v$ hX5e h;X5U0e[^H_Ë},1҅EE$@ ,݀>}E E $EEtfubBE,<}xXE}Et}E xS}E $EEtt11؅u"t-M0YEQ]0['hɋu0^US]3ZsS~ C t s XsSU]US]3oXsfXs]XsTXs KXsBYs9]]*UWVSj0}Ɖ$$F$F$F$ٹF $ιFE $轹FE F,9=h~ 19F}:V@9|19}NV@9|19~(}N V@9|F$e[^_Ã=hF ~v hXZ5lhVUWVSSE1҅ɋ}$u(]@B]DDy1ۉ`E MыEM A ЋEɋ AɉEEuE؋AMCMƋyY[^_ÉUWVS}O 11҃~$_9~B91~ W)4@9G [^_ÉUEÉUWVSM 1҅} uEx:l] Ƌ]B ˋy[^_ÍvUWVSPUH]$E,u<c8[@}4E(1ɅEE8<} }G}P$ɋE C ̋,݃uE4}EEE܋E(}݃͐EE؋ED}@vAC8EuEE^[^_EuEEKEcuC@EWuc@EBAIA:ɋM(K@ UWVSU11ɅExDu } ‹FHw(E} u+!X5E]311;E$[ t}@B;U$U$ R7uËN F1},ϋttA;M$|XUWVS,EU M]1=hEUM]EQu V農E؃1;}]@;E|Ep(11҃~6;M}1X ]Ћ]Ћ]uuEAp(B9~;M|ՋUJ EM~Dv1Ƀ~/EX vU؋tEPE@EAG9E]9]1ɋ};MG(}Bu^ A;M}-E؋tًMHUH]DGM@`؋}0F8EuEEe[^_ÉSvɋM0N@ M̋E$t4ȋMt4ًE̋M t4uuuurh0=h6U0Pr2hwE̋M$t4Et4؋M̋E t4uuuurh0=hnU0Pr2hMUUM1҃9vBP44u虝E@u1EكEu]C9|Ee[^_ÐC9؃$뿉UWVSLME }(U,MEM0E4}}8u ]UM܉E؉}ԋUKU0[uEEM4뾉UWVSP}19u]}A;M }:ɉMx,EETEuyEMyA9|~];M }VTEu5HEvx*U~KETEu4EMyאA;M |X[^_ÉUEM OvUEM 뚉USMt!E tA$Q[Ív1[ÍvUWVSɉMԉx0UԅEԋTEEԉMyA9|~e;M}^TEu=HEԍvx2UԅEԋTEEԉMyϐA;M|E U ‰1Ƀ~{;M}vu0EUuɉMԉx0UԅOEԋTE5EԉMyA9|~e;M}^TEu=HEԍvx2UԅEԋTEEԉMyϐA;M|u 19u}z‰~h;M}cU܋EU<E}0ҍ4}@ED}<ɉMԉx0EԅGEԋTE-EԉMyA9|~e;M}^TEu=HEԍvx2UԅEԋTEEԉMyϐA;M|E U ‰1Ƀ~{;M}vu0EUuɉMԉx0EԅWEԋTE=EԉMyA9|~e;M}^TEu=HEԍvx2Uԅ EԋTEEԉMyϐA;M|u 19u}z‰~h;M}cU܋EU<E}0ҍ4}DEH}<ډU6%}@lE$uu\$$hn$5萏k1E9}?} U B],u"} tU܋M$],щBU@9@9}EEfEl fE1m]m1;}E;}ɉMx6U'E](TE$EEMyA;M}z;M밋M8݁$,IB9cUU;U19MM@tB]8݃$,KBɋ]8݃K,܋$m]8݃K,܋$}8{Gz?_8WM$1} P_()Sw(+xV4D Puu - j.xU$ E119EL}}UU, uEM0u;E,݀U4,uu,=hCE@uK tu[^ÐFE@u!CFE@u[1^UWVS\EE؋}EW‹uG]؉]E]E ڋUԋr]ȅ<]t&v_tK9Nt0[uvu}EEM;M~4$~8뼐e[^à Ph8ҍvUWVS =hE ]E~ h7^56h"S*8ShAhP6 ]s$hW6 Mq(hW6Ur0r,hW6 Ep8p4hWi6uv@vtiB<:tt_B<:u hxm1Y[WhP2$2E Au e[^_1<:tVuhWX21@t1 7مۈt2DP t'Fu2DQ utF7uэSRh7P1E=]$9.7؅ۈt1ˋDK t3F܊&1DH utF7uэPRhSB1HuM1QPhR 1Hu]QPhR0Hu]S PRShQ0 u_~;~] 9D}^M ݅\C@ h| ho h. WWhZ h @UEu P/tà h.$/UWVSh]S/E  P.Eh~ h/Y5*.1]RShW.t;u }F W-t;u uF W.h~ h.X5-Ee[^_øD hL{-Xj. S h`-^Uj-t'UE UAEQAUEAQ à h-$6.UWVS UBu P_-]S1D@9~E0uN~M]SUE4^t1UrKJu؋[uG9}e[^_à h<,$`-UWVS U1: @]<Åu j p,u Pb,EE1ɉv]ˋT]ˉTA9|Et9 tu_u 1> @M <uXۉu P+tc19ى‰G |vW u  A9|]UGG_We[^_Ívo@KZ h*$,UWVS\5h]~ h+Y5*}WMQEPSTEhEE҉E~ h*X5u*Eu P*u Eu P*vUEu Pu*EQhS9+E|]Pu P2*Eh~ hp~*X5)MEv uQ)ZPuuu)@}?#tȃ }WuMQEP]SURuVuu0ME1]UĉދE܃T?}]T9~Ѝv=<M h?juuUȋ]̉U]E]Ћua$)MUHp Mxu=hUAPXMșduh)X55( u' u)Xu8)Xu/)5h~Puh5'MUe[^_øu h'Yj( h'[^uhU(_PuuhA( SF'_뻍vUWVShu~ h6 (_5/'tWpSlQVpPtpu PB'Dt P'@ h&hV'LVxVhL&As$RVhGL&{( WhGL&S, RhRLo& C4PhRLX&K< QhRLA&sD Vh]L*&{ WhGL& S RhGL% ShGL% CPhRL%u P%Cu P%`;CCǃDžHPLt@\%h$dpPXQTVPWCH QD@'0{h?jhD$$pH@@F@@ H93# L %X@-%XD!%=h~W IQh5#e[^_ð h#Xj$P@Hh_$Y߃ hH_#Yσ VS#UWVS }1Buu P#tkËt8vQE SK]EE@[tuwww VÉ4$$e[^_à h"$#vUWVS M0U 1ۅ vt.9Cu 9Cut 0Cuu P",UҋE 0t/9VQu؉SQ3uh?jhWbÉ<$"e[^_É}C0vC0}FSQr؋pVQuSvSQFAE@Yt h $!UWVS<} G1]Buu P!Ët8vQEȉSK]EE@[tuwww VÉ4$j!]Ћ(G1]B0uu P\ LËt8vQESK]EE@[tuwww VHE4$ XZ]uSۉEt* 3 C u{ s>$~ M u6c ]C uQ EpXu? }1ɺ?OWEЍe[^_à P t Ph?jh]SE[E벃 hv$UWVS M0U 1ۅ pvt.9Cu 9Cut 0Cuu PxUE 0t(tG9VQqu؋tvVQquSQ3uh?jhW*É<$e[^_É}C0vC0}SQVSQFiـE@Yt h$UVS]1t!sss 3jXsFe[^UWVS8hEQhuuNU P]MCu PX]11pC8ES8CU}OEMԋH}_T}TO]XP@1ۉU]t!sss 3BZsG jM]UԉX@MFE91XxWUtUBu PytH}UzG1Ƀ~v΋TTA9Ee[^_øk h$vUh$$$H$jÐUWVS] 1ɋ}uvϋTTA9[^_UWVSE E9EMUM U MU}uE}u\E\E 9E|E9E};Ev}uE}uM E\M9M\|σ[^_ÍvUWVS01;] E11;U }uΉB@;U |C}M ;] D?|EU 9ULUЋ] u!} ׋uEUЋ]JMu} u9|؋M IUY]] ٍ4 }41}B;U |؋U э4 ]D?quuxd] 4 M ىUMv1ɋU;M D}!UU }A,ǃ;M |u]MyM*؃0[^_ËuF} 9uuU׋]EЉ]MME ]ԉM؉U܉EȋM]ًEԋMЋU 9}"]]vUA,ƒ;M |1;] }!MMUC,ƒ;] |E̋]ȋM ]} u 9M|5UWVS1;]}}Wl1MʋU ɍ 19}M ًMB9|Cu;]|[^_ÉUWVS} E UU U1'ML]PDD@9Eu9uMMuu@uЋ]u}@U]uUEủEMEEUȉE1;}}O]u3UEM@4 MEHȋUGփ9|Ћ} 1;U}]uHM, B;U|E}u@E7EݕptEp]@L7 T7LTuE}OL19}711;U}%E,uHȍމ},B;U|A];M|ˋ]ԋM@ ݕptuȋp}u@]̋MĉGW TYUUEEUEuEE9Um19}KMD11;U}}H}(B;U|}8E4UCփ9|M ~[1;]}51;U}(E(uHM}, B;U|C;]|ˋM0U]D\gfffm ыu )|EEMEEMMMEEE}9}؋]]EE1҃1DžlEMD1Ƀ~!lE@v}PA9}@MP E5ًMLы}LMP$E ڋEMPlыEBu9eEEuU9U~u x1;MEL]uLD},ɋU,( }(ϋ]TA;M|1;]}t11;M}$}Hv}<ߋUTA;M|U<ڋM48؋}8(M<Cu;]|u ~؋U]LD#B ;1ۉpt;]݅p1M(ËU,1 9}}$U}< A9|;]}U$v}< A;M|U8M<Eu(EEuCu;]}E_ً}4EeEEtCu;]|1;M}*u4]<A;M|1ۋM9DžlE1}(E,1 9}U$ڋU}< A9|;]}U$v}< A;M|U8ڋM<% EuEEl}4 EJEEuDžlECu;](؋u ]01U<‹M, @;E|(EE5}XDžlE]}uEuuultEu M9M*U܅}9}DždDžl111;U}Al}Hd}<\ ]TBݕp;Upt|du<ދM4 ދUdl9di1u<@;E|M ~ELu(](ˋUTA9| hO v؋}Lً]8u U<AA؋d]8% u }<' ؍e[^_UWVSE]d4ËEE ]u؋ud ֍xSMԍ ڍ<U ։]Ћ]ۍCЉM }Ud<΍4E ‰}E}duUd4EM ؋]dM1E uĉEEE;M11ҋ}@;U}u<ސ}X B;U|UDʃ}EuT}LϋuXuEXUP"E}}LϋuXuEUPu$EuvA];M}ډMA];M|؃}}1UX<ʋt]ĉ<ˉtA;M|M9MEt } ] ~JMMM}<E4EUUUE0] /U uuu}<E4E M19M }V;E @;E|1;M}@1;Mt01;U})}< B;U};UuuB ΃;U|A];M|EuMuP<΋T]ĉ<ˉTuMuL<΋T]ĉ<ˉTEQMMM]<E4E/1;E|h@;E};EuU}@@;E]|1;M}D1;Mt41;U}-u< B;U};Uu} BE;U]|A];M|EEuM}P4ϋT]ĉ4ˉTEM}L4ϋT]ĉ4ˉTj[M}؃uuݝc^_uPEU݅E:ً]1Ƀ}<ϋTu؉ΉTA9} 4 }/1;}}t]]]uU D DG;}|DžD}9D}31]U؍4 DD] D}9D|ϋ} M9ME}:UщU}}E؋uu \UEE9E \|Dž@U9@}KDžTT}u؋T \BE؋u@T9@ Љ\|}0 1 }`@;E|u`u8u4uuuݝ}݅`U E9EE}:uu}U}UEЋuЋ\Ћ E\]9] |Dž8E98}KDž\\}UЋ\ \FEЋU8\98 \|u`]NFd 1;}}zUUxxE \ \G;}|Dž0}90}31UEЍ4 \0U \}90|ϋ}`d]]1Ƀv}<ϋTuΉTA9Eu61;]}/11vEDɍ< ЋuB;U|CM;]|1uHɋUD ‹M@]d@;E|؋} ~UHJ]K1;M}b1ېu`E U  $h5hu؋O~@t;kU$uEU}e[^_11ɋu}N~ hW 1]v_ E]QOt4hU1ۃ&vROt4hZC9P t4h^[Xp4hh Gt4hvFn97SQuuh~R} w7h=XZ5h+e[^_ÃjRtUB,k~H]C,OPh h61vQOt4hU1ۃ&vRGt4hZeC9ۃ G t4Ot4h:_t4hvF"97swPWr2h?~SOt4hC9PG pVShs P$$pNC؍ sэ($t111ҋM9QD@9|PhhUљGE3 hMd 1ɃEwUԉuċuċ΋TuԉΉTA9MyUWVS } EE9}E}MEv19}%UMѐPs3hF9| j CE}9}|E e[^_"UWVS1] 9ߋuE}1ɐ1 B9|G9|[^_ÉUWVS E E9E}BEEE EEv}uE}uU E \9U \|҃ [^_ÐUWVS [ÖBE)19s׉M)F9Ήr [^_ÉUWVS[EB) pNu. [^_ÐUSR vЋuX[US[APfY[ (more)14.08.08V6.02Arguments:General options:Learning options:Kernel options:Output options: More details in: 1999. 2002. trans_predictions Unrecognized option %s! be less than 1.0 !!! SVM-light %s: Support Vector Machine, learning module %s usage: svm_learn [options] example_file model_file example_file-> file with training data model_file -> file to store learned decision rule in -? -> this help -v [0..3] -> verbosity level (default 1) -z {c,r,p} -> select between classification (c), regression (r), and preference ranking (p) (default classification) -c float -> C: trade-off between training error and margin (default [avg. x*x]^-1) -w [0..] -> epsilon width of tube for regression (default 0.1) -j float -> Cost: cost-factor, by which training errors on positive examples outweight errors on negative examples (default 1) (see [4]) -b [0,1] -> use biased hyperplane (i.e. x*w+b>0) instead of unbiased hyperplane (i.e. x*w>0) (default 1) -i [0,1] -> remove inconsistent training examples and retrain (default 0)Performance estimation options: -x [0,1] -> compute leave-one-out estimates (default 0) (see [5]) -o ]0..2] -> value of rho for XiAlpha-estimator and for pruning leave-one-out computation (default 1.0) (see [2]) -k [0..100] -> search depth for extended XiAlpha-estimator (default 0)Transduction options (see [3]): -p [0..1] -> fraction of unlabeled examples to be classified into the positive class (default is the ratio of positive and negative examples in the training data) -t int -> type of kernel function: 0: linear (default) 1: polynomial (s a*b+c)^d 2: radial basis function exp(-gamma ||a-b||^2) 3: sigmoid tanh(s a*b + c) 4: user defined kernel from kernel.h -d int -> parameter d in polynomial kernel -g float -> parameter gamma in rbf kernel -s float -> parameter s in sigmoid/poly kernel -r float -> parameter c in sigmoid/poly kernel -u string -> parameter of user defined kernelOptimization options (see [1]): -q [2..] -> maximum size of QP-subproblems (default 10) -n [2..q] -> number of new variables entering the working set in each iteration (default n = q). Set n size of cache for kernel evaluations in MB (default 40) The larger the faster... -e float -> eps: Allow that error for termination criterion [y [w*x+b] - 1] >= eps (default 0.001) -y [0,1] -> restart the optimization from alpha values in file specified by -a option. (default 0) -h [5..] -> number of iterations a variable needs to be optimal before considered for shrinking (default 100) -f [0,1] -> do final optimality check for variables removed by shrinking. Although this test is usually positive, there is no guarantee that the optimum was found if the test is omitted. (default 1) -y string -> if option is given, reads alphas from file with given and uses them as starting point. (default 'disabled') -# int -> terminate optimization, if no progress after this number of iterations. (default 100000) -l string -> file to write predicted labels of unlabeled examples into after transductive learning -a string -> write all alphas to this file after learning (in the same order as in the training set)[1] T. Joachims, Making Large-Scale SVM Learning Practical. Advances in Kernel Methods - Support Vector Learning, B. Schlkopf and C. Burges and A. Smola (ed.), MIT Press, 1999.[2] T. Joachims, Estimating the Generalization performance of an SVM Efficiently. International Conference on Machine Learning (ICML), 2000.[3] T. Joachims, Transductive Inference for Text Classification using Support Vector Machines. International Conference on Machine Learning (ICML),[4] K. Morik, P. Brockhausen, and T. Joachims, Combining statistical learning with a knowledge-based approach - A case study in intensive care monitoring. International Conference on Machine Learning (ICML), 1999.[5] T. Joachims, Learning to Classify Text Using Support Vector Machines: Methods, Theory, and Algorithms. Dissertation, Kluwer, The parameter rho for xi/alpha-estimates and leave-one-out pruning mustbe greater than zero (typically 1.0 or 2.0, see T. Joachims, Estimating theGeneralization Performance of an SVM Efficiently, ICML, 2000.)! Not enough input parameters! Unknown type '%s': Valid types are 'c' (classification), 'r' regession, and 'p' preference ranking. It does not make sense to skip the final optimality check for linear kernels. Maximum size of QP-subproblems not in valid range: %ld [2..] Maximum size of QP-subproblems [%ld] must be larger than the number of new variables [%ld] entering the working set in each iteration. Maximum number of iterations for shrinking not in valid range: %ld [1,..] The fraction of unlabeled examples to classify as positives must The C parameter must be greater than zero! The COSTRATIO parameter must be greater than zero! The epsilon parameter must be greater than zero! The parameter depth for ext. xi/alpha-estimates must be in [0..100] (zerofor switching to the conventional xa/estimates described in T. Joachims,Estimating the Generalization Performance of an SVM Efficiently, ICML, 2000.) It is necessary to do the final optimality check when removing inconsistent examples.|ɓbC#ʒ۔}^? yZ;Calculating model...done Cache-size in rows = %ld Kernel evals so far: %ld done. Shrinking... Reorganizing cache...inconsistent(%ld).. (joint %f) (single %f) done %ld..Running optimizer...(i-step) Selecting working set... Iteration %ld: %ld vectors chosen (j-step on %ld)Writing alpha file...w%.18g Writing prediction file...%.8g:+1 %.8g:-1 %.8g:-1 %.8g:+1 pos ratio = %f (%f). Retraining.Number of switches: %ld Computing starting state...L1 loss: loss=%.5f Runtime in cpu-seconds: %.2f done. (%ld iterations) OptimizingNumber of SV: %ld Computing leave-one-out Retrain on full problem(?[%ld]-)+)?BOGO?ffffff?333333?dAY@?333333?-C6? Number of inactive variables = %ld sid %ld: dist=%.2f, target=%.2f, slack=%.2f, a=%f, alphaslack=%f Computing qp-matrices (type %ld kernel [degree %ld, rbf_gamma %f, coef_lin %f, coef_const %f])...Error: Kernel cache full! => increase cache sizeObjective function (over active variables): %.16f WARNING: Relaxing KT-Conditions due to slow progress! Terminating! Checking optimality of inactive variables... => (%ld SV (incl. %ld SV at u-bound), max violation=%.5f) xacrit>=1: labeledpos=%.5f labeledneg=%.5f default=%.5f xacrit>=1: unlabelpos=%.5f unlabelneg=%.5f xacrit>=1: labeled=%.5f unlabled=%.5f all=%.5f xacritsum: labeled=%.5f unlabled=%.5f all=%.5f r_delta_sq=%.5f xisum=%.5f asum=%.5f POS=%ld, ORGPOS=%ld, ORGNEG=%ld POS=%ld, NEWPOS=%ld, NEWNEG=%ld Increasing influence of unlabeled examples to %f%% .Model-length = %f (%f), loss = %f, objective = %f %ld positive -> Switching labels of %ld POS / %ld NEG unlabeled examples.Classifying unlabeled data as %ld POS / %ld NEG. %ld positive -> Added %ld POS / %ld NEG unlabeled examples. Moving training errors to inconsistent examples... Now %ld inconsistent examples. Setting default regularization parameter C=%.4f Optimization finished (maxdiff=%.5f). Runtime in cpu-seconds: %.2f (%.2f%% for kernel/%.2f%% for optimizer/%.2f%% for final/%.2f%% for update/%.2f%% for model/%.2f%% for check/%.2f%% for select) Norm of weight vector: |w|=%.5f Number of SV: %ld (including %ld at upper bound) Norm of longest example vector: |x|=%.5f Number of kernel evaluations: %ld 'remove inconsistent' not available in this mode. Switching option off!Number of non-zero slack variables: %ld (out of %ld) Error: Missing shared slacks definitions in some of the examples.Number of SV: %ld (plus %ld inconsistent examples) WARNING: Using a kernel cache for linear case will slow optimization down! Deactivating Shrinking due to an incompatibility with the transductive learner in the current version. Cannot compute leave-one-out estimates for transductive learner. Optimization finished (%ld misclassified, maxdiff=%.5f). Estimated VCdim of classifier: VCdim<=%.5f Computing XiAlpha-estimates...Runtime for XiAlpha-estimates in cpu-seconds: %.2f XiAlpha-estimate of the error: error<=%.2f%% (rho=%.2f,depth=%ld) XiAlpha-estimate of the recall: recall=>%.2f%% (rho=%.2f,depth=%ld) XiAlpha-estimate of the precision: precision=>%.2f%% (rho=%.2f,depth=%ld) Cannot compute leave-one-out estimates when removing inconsistent examples. Cannot compute leave-one-out with only one example in one class. %ld positive, %ld negative, and %ld unlabeled examples. Leave-one-out estimate of the error: error=%.2f%% Leave-one-out estimate of the recall: recall=%.2f%% Leave-one-out estimate of the precision: precision=%.2f%% Actual leave-one-outs computed: %ld (rho=%.2f) Runtime for leave-one-out in cpu-seconds: %.2f Leave-One-Out test on example %ld Constructing %ld rank constraints...Error: Unknown kernel function%ld # number of training documents %ld # number of support vectors plus 1 %.8g # threshold b, each following line is a SV (starting with alpha*y) Line must start with label or 0!!! Feature numbers must be larger or equal to 1!!! Slack-id must be greater or equal to 1!!! Cannot parse feature/value pair!!! Features must be in increasing order!!! Not enough values in alpha file!Reading examples into memory... Maximum feature number exceeds limit defined in MAXFEATNUM! Parsing error in line %ld! %sOK. (%d support vectors read) Parsing error while reading model file in SV %ld! %sVersion of model-file does not match version of svm_classify! Copyright: Thorsten Joachims, thorsten@joachims.org This software is available for non-commercial use only. It must notbe modified and distributed without prior permission of the author.The author is not responsible for implications from the use of thistJJJ(KhJWriting model file...SVM-light Version %s %ld # kernel type %ld # kernel parameter -d %.8g # kernel parameter -g %.8g # kernel parameter -s %.8g # kernel parameter -r %s# kernel parameter -u %ld # highest feature index %.32g %ld:%.8g #%s r%lfqid:%ld%ssid:%ld%scost:%lf%s%ld:%lf%s'%s' in LINE: %s Out of memory! Reading alphas...%lf Scanning examples...OK. (%ld examples read) Reading model...%ld%*[^ ] %lf%*[^ ] %[^#]%*[^ ] software. $tI#tI ASOLVE DUAL: inappropriate number of eq-constrains!before(%.30f)...after(%.30f)...result_sd(%d)... WARNING: Relaxing epsilon on KT-Conditions (%f). V瞯- ? ?real_qp_size(%ld)...%f: %f : a=%.30f: nonopti=%ld: y=%f eq-constraint=%.30f b=%f smallroundcount=%ld return(%d)...: a=%.10f < %fEQ: %f*x0 + %f*x%ld = %f return_srd(%d)...%5.2f @ D HP  L \Dooor‡҇"2BRbrˆ҈"2:0yE>h㈵>GCC: (GNU) 3.4.6 20060404 (Red Hat 3.4.6-9)GCC: (GNU) 3.4.6 20060404 (Red Hat 3.4.6-9)GCC: (GNU) 3.4.6 20060404 (Red Hat 3.4.6-10)GCC: (GNU) 3.4.6 20060404 (Red Hat 3.4.6-10)GCC: (GNU) 3.4.6 20060404 (Red Hat 3.4.6-10)GCC: (GNU) 3.4.6 20060404 (Red Hat 3.4.6-10)GCC: (GNU) 3.4.6 20060404 (Red Hat 3.4.6-10)GCC: (GNU) 3.4.6 20060404 (Red Hat 3.4.6-10)GCC: (GNU) 3.4.6 20060404 (Red Hat 3.4.6-9).symtab.strtab.shstrtab.interp.note.ABI-tag.hash.dynsym.dynstr.gnu.version.gnu.version_r.rel.dyn.rel.plt.init.text.fini.rodata.eh_frame.ctors.dtors.jcr.dynamic.got.got.plt.data.bss.comment#(( 1HH7   0?PPLGoFTo`c DDl \\ uDDp\\{<< ((0JJJJJJJppK(K K3MlR 2 L` (H PD\ D \ <  (p` *8ExIU k w   \V 5CKIJ cpnVpn Y   ,~ $i 2 K`|PE l ~R l= M l<  t  Pm \[` )e 0X: :l, K]h {    0~ A I# J TT Q M h x  1    @ x O(   B  , \ ; {< S N b \'  3 S9  p X  , hL+ $ Dm4 1 C 0ZV R p_ s z   "[  ,T( \ p f^   Z ?  K Z C l \1  0call_gmon_startcrtstuff.c__CTOR_LIST____DTOR_LIST____JCR_LIST__p.0completed.1__do_global_dtors_auxframe_dummy__CTOR_END____DTOR_END____FRAME_END____JCR_END____do_global_ctors_auxsvm_learn_main.csvm_learn.cswitchsens.0switchsensorg.1switchnum.2svm_common.csvm_hideo.ckernel_cache_cleanupopt_precisiondocfileclassify_example_linearnonoptimal__strtod_internal@@GLIBC_2.0calculate_qp_objectivemodelfilefeof@@GLIBC_2.0read_alphas_DYNAMICsingle_kernelcustom_kernelroundnumberfree_modelladd_matrixsolve_dualappend_svector_listread_input_parameterskernel_cache_space_availablecache_multiple_kernel_rowsselect_next_qp_slackset_fp_hwperror@@GLIBC_2.0fprintf@@GLIBC_2.0fflush@@GLIBC_2.0clear_vector_n__fini_array_endadd_to_indexclock@@GLIBC_2.0minlmaxitercopy_svectorverbosityselect_next_qp_subproblem_gradprint_help__dso_handleread_modellindep_sensitivity__libc_csu_finilswitch_rows_matrixcopy_modelputchar@@GLIBC_2.0identify_one_misclassifiedcopyright_noticepow@@GLIBC_2.0estimate_margin_vcdimisnan@@GLIBC_2.0select_top_ninit_shrink_statelswitchrk_matrixsmallroundcountparse_documentputs@@GLIBC_2.0_initdistribute_alpha_t_greedilyfeatvec_eqmalloc@@GLIBC_2.0maxlfscanf@@GLIBC_2.0kernel_cache_reset_lrucache_kernel_rowlprint_matrixsub_ssstdout@@GLIBC_2.0estimate_r_delta_averageprecision_violationswrite_modelkernel_cache_freeupdate_linear_componentoptimize_to_convergencecompute_xa_estimateskernel_cache_statistic_startcalculate_svm_modelfgets@@GLIBC_2.0bufferfree_examplecreate_svectoradd_ssmy_malloccheck_optimality__strtol_internal@@GLIBC_2.0kernel_cache_checksmult_slinvert_matrixsvm_learn_classificationoptimize_qpcompute_matrices_for_optimizationwrite_predictionoptimize_svm__fini_array_startidentify_inconsistent__libc_csu_initincorporate_unlabeled_examplessprod_sscompute_objective_functioncompute_shared_slacks__bss_startreactivate_inactive_examplesmainkernel_cache_malloc__libc_start_main@@GLIBC_2.0__init_array_endexp@@GLIBC_2.0free_svectortanh@@GLIBC_2.0wait_any_keycheck_optimality_sharedslackkernel_cache_clean_and_mallocdata_startprintf@@GLIBC_2.0_finimemcpy@@GLIBC_2.0sqrt@@GLIBC_2.0write_alphasfclose@@GLIBC_2.1kernel__preinit_array_endkernel_cache_shrinkshrink_problemestimate_spherelength_of_longest_document_vectorsprod_nsnol_llsvm_learn_optimizationexit@@GLIBC_2.0restartfilesscanf@@GLIBC_2.0_edata_GLOBAL_OFFSET_TABLE_free@@GLIBC_2.0compute_index_endadd_vector_nskernel_cache_free_lruget_kernel_rowkernel_cache_initstdin@@GLIBC_2.0read_documentsoptimize_hildreth_despomodel_length_sestimate_transduction_qualityselect_next_qp_subproblem_randfopen@@GLIBC_2.1get_runtime__init_array_startestimate_r_deltaoptimize_to_convergence_sharedslack_IO_stdin_usedclassify_examplelcopy_matrixfwrite@@GLIBC_2.0create_example__data_start_IO_getc@@GLIBC_2.0_Jv_RegisterClasses__ctype_b_loc@@GLIBC_2.3__preinit_array_startdualsvm_learn_regressionspace_or_nullshrink_state_cleanupidentify_misclassifiedadd_list_ssprimaladd_weight_vector_to_linear_modelclear_index__gmon_start__svm_learn_rankingkernel_cache_touchstrcpy@@GLIBC_2.0