ai4r 1.3 → 1.4

data/site/build/site/en/linkmap.html

@@ -115,13 +115,13 @@ document.write("Last Published: " + document.lastModified);
 <a href="http://forum.jadeferret.com/viewforum.php?f=3" title="Ai4r Forum">Forum</a>
 </div>
 <div class="menuitem">
-<a href="http://ai4r.rubyforge.org/rdoc/index.html" title="Ai4 RDoc docs">API doc (Rdoc)</a>
+<a href="svn.html" title="ai4r Subversion repository">Source Code repository</a>
 </div>
 <div class="menuitem">
-<a href="svn.html" title="ai4r Subversion repository">Source Code repository</a>
+<a href="http://wiki.jadeferret.com/Category:AI4R" title="ai4r wiki with How-to guides and more documentation">AI4R Wiki</a>
 </div>
 <div class="menuitem">
-<a href="wholesite.pdf">ai4r project PDF</a>
+<a href="http://ai4r.rubyforge.org/rdoc/index.html" title="Ai4 RDoc docs">API doc (Rdoc)</a>
 </div>
 </div>
 <div id="credit">
@@ -208,23 +208,21 @@ google_ad_height = 125;
     
 <ul>
 <li>
-<a href="http://ai4r.rubyforge.org/rdoc/index.html">API doc (Rdoc)</a>&nbsp;&nbsp;___________________&nbsp;&nbsp;<em>rdoc</em>&nbsp;: Ai4 RDoc docs</li>
+<a href="svn.html">Source Code repository</a>&nbsp;&nbsp;___________________&nbsp;&nbsp;<em>svn</em>&nbsp;: ai4r Subversion repository</li>
 </ul>
     
 <ul>
 <li>
-<a href="svn.html">Source Code repository</a>&nbsp;&nbsp;___________________&nbsp;&nbsp;<em>svn</em>&nbsp;: ai4r Subversion repository</li>
+<a class="external" href="http://wiki.jadeferret.com/Category:AI4R">AI4R Wiki</a>&nbsp;&nbsp;___________________&nbsp;&nbsp;<em>wiki</em>&nbsp;: ai4r wiki with How-to guides and more documentation</li>
 </ul>
     
 <ul>
 <li>
-<a href="wholesite.pdf">ai4r project PDF</a>&nbsp;&nbsp;___________________&nbsp;&nbsp;<em>ai4r_site_pdf</em>
-</li>
-</ul>
+<a href="http://ai4r.rubyforge.org/rdoc/index.html">API doc (Rdoc)</a>&nbsp;&nbsp;___________________&nbsp;&nbsp;<em>rdoc</em>&nbsp;: Ai4 RDoc docs</li>
+</ul>    
   
 </ul>
 </ul>
-  
 
 </ul>
 </ul>

data/site/build/site/en/linkmap.pdf

@@ -5,10 +5,10 @@
 /Producer (FOP 0.20.5) >>
 endobj
 5 0 obj
-<< /Length 1145 /Filter [ /ASCII85Decode /FlateDecode ]
+<< /Length 1206 /Filter [ /ASCII85Decode /FlateDecode ]
  >>
 stream
-Gat=+fi$QY'Re;/p^q+qX2XYH>9K^N6Q&MeeoVbL<ZDjp8nEe>mjm-a,ag&J9kUE-N:&_#kUuSZD")S+*TXcHR4SZ:r\#,h-Q>;UaFDQnl94#,0(qYF[W[+j9j'p2SI6-bo#]Q*/"YA_7W$B/7BSf:9=;Lr77E-74El69ZWi8[*/Id/"-*M8lN=>ES#odq]--d11OU:Ejc2WB[#j[d>#AiVI(dX9V<sL@\)JU^O=C^>XMd#r[Xu'qma/hA6uE_,\Bba2-\@&FQA,+Lp%HeQbS(VTJ-P`3j3PQ4s1o^"aRsu1?Q5Z/!V1"1*cquJ3+"r'7fJs^Bl?=8jC/glro41><i7AAo`d_Q[tW8+8Ej>c:)IK_aYNs"ld<i]926?#k+PF"ES=XO=Vkg5'GXiXKZ1LUS&cepF-b2td;j0I`K%`"Lp4uKR3UZ3no;9dn,?<3QPW,k@QJNmLcDA9U/SRIMsEt`H]&_pN?/;jAri_"gFkRe1Z&SI9^T(`7LL]ZBi=(2(T^fBHaEFMmm[mE@C$_>pMOP::`sO81`]hE/d0#%Uc]9gPWp23^>BW>/WSV-Wu)>KAaq!?<ijYGlAilY?q&XR0A6"aBS*Bk']n)P(6iS0S`:Knj8_4dq.,.Y$mV4FP&X3cR@s2eAHTkS*qsb,o\-gQeLMH^rNf2;d\s$Q((H(?D&3Xh2[Xs(oPr8tKN&^n`Q@^@n4fFlPp)J?][$u<oG8KTl\$,Ff-9Btg:s,uJL0KFJW]AjSrh*IYflAh/U;I-]\aS_eZ:f?'@0j"@kP4Ur_obt#m4SgH9<p9A@E#0Gt8<gBS8M!]QKsc>.&LI)EpiZMj/!?COYt*mJ2%rO@.[5(`DP3JZrK9O<$D>O-fquj)Lm$Q$t;d$LOs'b-s^0ak&Z*/3u4"^.r\<hRbY\W#;'/L[q9$V#[YG/M`AhdsS`4]MR#Lb9QHg(hapW&*JaD.(!'g#qF]T+.?/Dj'ae4T-U3Md0=n+pJeoK5&Ja0mdSJ;n@/3gBjS@o#i'sOTj&JO7Doef=UoljY3PHp,bOP?DaG._7/OWJEg0';E?ab4TS:#?Rs2]S<M_lak(l96+TMRT9rZ2N=8&"Y6&Y;lg:<2Am=T(GKqOn'@@f!E_(@O6Q;kNDe5]rW"+`m"o)~>
+Gat=+gMYb*&:O:SkV3)r>$aU&qW>`d`^R@M'*J-Y'FR@(AKm<$V5RYu^6E/tUgb6JV*N1s_Du6=B/nuFHM6M7HRSB7\),=opKof\?87'R;#d2TFE:jjHlU/_](1;T4>9>[]djf9lH@-_<^0[M:/@t?Vpb@c2gTkhCIr<p>^AWJok/tP4GX2P$][W;++dmN>AmZn\KK^p$Zhj'iL[kn%$8A/DQ,NL?L:?7NL>Nh<a+YjOu/gM.G@J(K)/"*d(u7uIM2+[59`M@":ILC$OMA+*26/1+;?@-4m]p*Iu`s6&+0upIMgf<L&g$<:tRK.TT=WBq$Ot^niL1*ab5hqa!eCpVbasjk9u&?"@;=^K@6dFl.]04(h:N6YQ<D7dE)\F:)J`ZU/+8<%(isWY4QZc,hoP^V59bfW?N``<mWl`Ec8k7T"n3Z\BcOf,Okg*^3t;Ls&C^R&1>%u7WNnjBdQtp;Rh-J]1Q``MjVGr*Z",1J>9fi2`^+X5DY7l*O5U%el6oNUaum+R+t`+=U-qYP6BRC`IMHEGHrPD;RIP"%R,k1"Mq7:!YXa^59J)-]W(R+SCm!j0cW9uipS_YZJ%jK0/OD$X,H4gWDanW2J=t@B]!^8TMd9[/U+.i)UL4\<su*>=O<!aN#UKP"f;6'2F*eonWgQ"XGsD_^2ia+4j-eU]kSm+h1KCmDUAIXP:3e>T,%k`88$n(RQ^*Bhk`ak/\)k[rP!g\q19jXljsR(>*h@rVk5<9@B$tP@H;d5B;>kW*&IJ_,C+;$I$e%lI9;KG-@rMYQ<^fIoI_uMKE\Rt4J(e(lElP4]uUJ)DM>_[]T&Z&7^\5MHJknZgX6PW%85%g=#UI=hBuA09`!%CEG/\DF8Lp.AqTFrcCKDtT3eH5Nf4Y='I0RBapdHt:&m</dsMncCX@R`1RjmAlG=BmJW9BW4*5+%gthU3H]:se_3qam<a0eGoh'7A(G3fp(.0_q;U&P*ilnHD`J'F"3R?DO?1[dqd0X6DlCY5bRX\OEY-d_DENe`LbK3)kYeQT1kb<iOk8mH$L=fgA=.D2:(FEAJpE8M`HM1_GVWo6/&4Ct8e-2O%J<,L6W4/1CT3ZYZ0r2P\qmT[dU4=(V0rS2"1%>nCPA&lp_ftR$'jIXZ=c\A(D'8m'D*:]m0WsdhUOq(A++!'jRN%<s,o)"M^?9(,N?5P`Ct__U6??Lhr<lH0??l~>
 endstream
 endobj
 6 0 obj
@@ -72,17 +72,17 @@ endobj
 xref
 0 12
 0000000000 65535 f 
-0000001969 00000 n 
-0000002027 00000 n 
-0000002077 00000 n 
+0000002030 00000 n 
+0000002088 00000 n 
+0000002138 00000 n 
 0000000015 00000 n 
 0000000071 00000 n 
-0000001308 00000 n 
-0000001414 00000 n 
-0000001526 00000 n 
-0000001635 00000 n 
-0000001745 00000 n 
-0000001853 00000 n 
+0000001369 00000 n 
+0000001475 00000 n 
+0000001587 00000 n 
+0000001696 00000 n 
+0000001806 00000 n 
+0000001914 00000 n 
 trailer
 <<
 /Size 12
@@ -90,5 +90,5 @@ trailer
 /Info 4 0 R
 >>
 startxref
-2197
+2258
 %%EOF

data/site/build/site/en/machineLearning.html

@@ -115,13 +115,13 @@ document.write("Last Published: " + document.lastModified);
 <a href="http://forum.jadeferret.com/viewforum.php?f=3" title="Ai4r Forum">Forum</a>
 </div>
 <div class="menuitem">
-<a href="http://ai4r.rubyforge.org/rdoc/index.html" title="Ai4 RDoc docs">API doc (Rdoc)</a>
+<a href="svn.html" title="ai4r Subversion repository">Source Code repository</a>
 </div>
 <div class="menuitem">
-<a href="svn.html" title="ai4r Subversion repository">Source Code repository</a>
+<a href="http://wiki.jadeferret.com/Category:AI4R" title="ai4r wiki with How-to guides and more documentation">AI4R Wiki</a>
 </div>
 <div class="menuitem">
-<a href="wholesite.pdf">ai4r project PDF</a>
+<a href="http://ai4r.rubyforge.org/rdoc/index.html" title="Ai4 RDoc docs">API doc (Rdoc)</a>
 </div>
 </div>
 <div id="credit">
@@ -229,7 +229,7 @@ The only information that you have is a collection of examples, provided by a ma
 	and identify new posible marketing targets:</p>
 <pre class="code">
 
-  id3.to_s
+  id3.get_rules
     # =&gt;  if age_range=='&lt;30' then marketing_target='Y'
           elsif age_range=='[30-50)' and city=='Chicago' then marketing_target='Y'
           elsif age_range=='[30-50)' and city=='New York' then marketing_target='N'
@@ -277,7 +277,8 @@ id3.eval(['New York', '&lt;30', 'M'])
     # =&gt;  'Y'
 </pre>
 <p>
-But instead of going through the tree every time, you can take advantage of the fact that the method "to_s" generates proper ruby code!
+But instead of going through the tree every time, you can take advantage of the 
+fact that the method "get_rules" generates proper ruby code!
 	</p>
 <pre class="code">
 
@@ -286,7 +287,7 @@ But instead of going through the tree every time, you can take advantage of the
     city = 'New York'
     gender = 'M'
     marketing_target = nil
-    eval id3.to_s
+    eval id3.get_rules
     puts marketing_target
       # =&gt;  'Y'
 </pre>

data/site/build/site/en/machineLearning.pdf

@@ -95,10 +95,10 @@ endobj
 >>
 endobj
 20 0 obj
-<< /Length 1797 /Filter [ /ASCII85Decode /FlateDecode ]
+<< /Length 1802 /Filter [ /ASCII85Decode /FlateDecode ]
  >>
 stream
-Gau0DD/\/e&H;*)Tqrjt!!t0=H7b9]8X^>FdUGk.!C[tFN()JfD98R:AAPkW^=+T.M:?cPL*DnCV!e2Q48O%]3$/\1(f4uKIieoh9l/#E4S[fRoCkW;qS`+Z-*=eqDhHO(]M1N/[p@0(534uOT0B>YQ<IDhRo^]6>pR(r#VDK`NTUDU:s=ZD'\6Cc/ItAh>3!#51Q!DPF?7\?X]oT=fc6e_IZT8:o((_.II#7dF.T/ep[Cp\8]5\V+Vo*u7'_opa29->&'^,@b/@+lTtX/?-]=r>C*E9?nD!H3!ek2,%12W!"lq0!PBFe16e@pr,Cq%q$o*\5,9/dL`\XkjT;g?R>^G!3M*jqH4"KWO+`Z>i;YI/=0PfQDomI(Qh&mtLfVeGp*jts."T2i=?2H=[DuW7"qrE#G$"&_Wqqr4[a7X78Bd67h(.uN4*0-@`+o4K8L.G3Z#aqLC`R<+Q1/HG(kVu!<7U$&M+n\-+dU#L,!si):.YXT23QA^L<4\2!bE'*<[WG=OHS.db6p`,)E[B<md6h_Uo[K!,A@UDA/iSk8n&9#s]'7;a"6@rO"#0[bGWXUI6lMBmjS;#1)AD0=5QaUZp#,'hhgoF]67COp5^PC.!-LVh0ZP:a!1hS*:u<`Z!sEIpa9c+gJTIM7-_MP3=H$"OejWsr\oMp^X\o`G$"+0TFUZ=b$+7H9+3rhreT!1`]%$70oS?D7:Pk43Uf\f)nQ[!t1#3c-+^UjpgC0WWMrYIIWKM\\U&;&ojI]np=ip"19@D!IG!]sZhefdFO@RXF,`AGQVp?k>pQJf+FF)i-%JKJ7+g8LIBS,o2i$I#gB@jm;@PXCKlS[[cac0+`,86^o[PH%t]3^oG.nSGDqk&2Kj"TJ^*!l_qfCRb9lTfW?@?n./4/SE6Lj!T@Df.Y;hm$I#&JYE;bI_t";F"Uqp`Pcad7%q$^eup*mCi=rTV<:9\V@f^o]"?]q/e[Ns8-XbS'828]n<Jlo$E*Q:e;S*iFK]BBT6=(&q?kPiC>8]nA#.Qmksk'#^(^jU!mBtE5:Fi73<js(r:I?XXeAr?)H)WNTo+OT'L9CNX$Eb2'CYqkL0Z2')=4q(`kf%;<lNDlt\j4kH'A34(BRL'JPi<(-/>!:X&,K9d<W/nHRN>g0;RBs4LFI[g,bji+31:<,3KS:^4Ns.a."b3Y!h[(ocG@6d/RTb7"]*q&[<YXWVc<;D26dVU%q%@ZmIb,.h1q6s)ugX`4`(6.^A_7EY']#OWGqru2!+l]BJ*]?UETiQH-u2-?De[;fcJA/]F1/>R?WLTMVlQ`?MGRKXBKXe`lIPj/kDn62QpfBTe#giKjtNa#n50+n^=E4:9Jq/\:>GY@6J#G34k@kR:e[lhpU^&/5ILR]M$R<Q,`IW+_#UUGBfk#iVE?X>N]W=M:2Z8;iCs&,/K+A0Nk^Y"PhqVbt41;,jnUC8`?/l,6g0<]^8Pm3K*FX8kGQ;0`r/\$^q&("EkH[b)Brk@\6fYu+1K\&Hs_2Idl!V?3:o=[eR'<pno@je0"??L=0..M'Pk7E2^.R3g0JT>M`A[BpcMa#aj!WZg9`-GD`OA^b;l5':dba*6H<"9>h@4ku1.>=dWeRl'-:7HqElPo\QSAeWVbqb2e^k&Sp'"8BV'ifk.!]j;S-+/0\]2a:rQTiLYH(Ecue$@=DJR9;gCBYE3,`^[.r:E>XkiaYu!Q^&e#[Am!^dlrMI$SXk<P2EL(ube>\BA,l.)90H:tJSsimc<k1#/:_EmQI@5/V2mp8(Y"nNJDj)^SA&bn)H_WU=>N3o3*/qK]m~>
+Gau0DD/\/e&H;*)Tqrjt!!t0=H7b9]8X^>FdUGk.!C[tFN()JfD98X<AAPkW^=+T.M6qM0L*DnCV!e2Q48O%]3$/Z[%&gRa^Q4#o-FW#]*rPK:q>2=XIH#U>P1JII\%ud$hBniSgTG*$T5pRc:ENUg9.d7Ec)ZoV/e[#IKGMhA`q+ae-t^@]$>V3mQ3SZDXmt,V)9!18$`U%=]9:j]F_ct$S(#m/:U?-Z?ClVPD]O.Ic?OI/*o8=rM0Nnr`'R4a8/CdkG^.0oU"RgdR2BZje9@nPA22V$JI9H8Do8I[']sZt,g7nVOK5HYed])!?$il<"H"H[`%9K#E9qBYcbq#UVH-?/l!c_"%P?f1XK3@q`-#E[frst_XF3alpDqb,K91=dTA!Y!eFl6Bn<_[KmDH`Rr>.3UPNRWFHo\(mK3-,9f<Yd[N@\dG-kTs03*o42kQd4f,"]0cA<LgH91-bA4B60jitNpZD&$`>e-BtbHR8%f+DB$"'T,fL)"Y=9ksd-R5MFE:dQo&C\JU_I$&s2OQ!DL#&$^9);b%V5iGtEtgu*q"(TY[AVB\_EOAg3/_"<uPm:<sFBj:kPkPm?dQloHY)Ggdf#>^EV%Qm)o6IAg^g9]J*W"XK13<hqRkaWEIqGs:-^h>5d:>c`pS-r=W!W?-:@WX*-/cnu!Q)cEMQ&;M2!+S9pD)f7*W9lJ=E"3D[Nb_MCV2iTi;3]d^8rPkY`<o1tkL*1fl.-CVnHp;lB'A#R,`daRQ0bgej`e*Vrt+K*oT]M08MBtn$*@3T\(*6$*^*/[/<4bC5-L4+:Mu0=l$Z2_f[h't]^?]3$.]oEQXHGV?2neua_7!Cf8*3L2aRh19Rr#Qat1m-H=\/$,\a1CIqnnA_h`5O,Rd4Nj)5a;"53'q<QaIlQt?sSg9Gq:\a?JIa*+.F<"-Pqn(q>A!J)EPK1o$0(V-4>9?C_OpZBeRMIQ);)YU&WMbf]&e*1ECYICoU1<]+[%_BO"6Pl1Y-#'<4PPr@l20pE:r2J,"T:K*$!tRZ3fA*,Ge-49U:4o7U!J`0@I@/)a%c].MY/D?&R82'dVJ"(sY.OKi*akYmJ8?U*:EMPfjMRd%Gjn*jWC$coY5soYZPGIdN0To1^fXW@jIjXZ0s?KlNDOo?)cT%9D].a:B3R7L*^o+?QQajUX`XV/)=.F(/>(tf35Ad,Uk81I0jBJc1F2VSHN5S'K%+(!]kj+)4u3A!8W'C&m!_Y?@3L?4`[r/7?0;A*.DPHo\7i,HDZjIN6UI:03Wo#00>HMr@!0'2`g(P,Btu=,ONqhe[X*a9WbTgp(/T[kOqL8]W^iSsCPgbOX4tp=cdC`6DLp3K82Q:@h>4P?cc&T</YSdVI#b:e\\t]-nh_H_RX<s6nq]tMhW!6\^;2Ucg>4!jB5&W@H-K<b5")\JeK?"69;u)=.1L>0@lkn!LQh8Mo_Nrod^3,f4?5R>]^EmB)Gjnpjj;de3X&Mbi@ZH@i+&>Kq7X!/qlZu-(H7Bf?DB(T0>c<DbjS3NpF[ae^q-_1SXbjfe]l40`2ij/irlrr]_J6(jjX:TAuecnIBHNQ"j@=0o*bpJY6\^iIJA_f/`WY*+W.W)>f,]&$<<.hc!ii52@C8b2tR>8MmMukAJW=:#[(phDBHA<8d3`LoVt5NCAYqnX!B@$Iur[]3lPg2AYDHm^G>0-$8NMk*4_pA-&;]NgKj`(O:O2]\5qO<O/bMB/DWFgGd/gH`".aAQatUBJAT$oY,,=?WM(tS\3hY*aBu)tL+*`%9;NonI8\oa8tqE:+^/JUh9-)s^:s]'id_"~>
 endstream
 endobj
 21 0 obj
@@ -288,44 +288,44 @@ endobj
 xref
 0 39
 0000000000 65535 f 
-0000009543 00000 n 
-0000009622 00000 n 
-0000009714 00000 n 
+0000009548 00000 n 
+0000009627 00000 n 
+0000009719 00000 n 
 0000000015 00000 n 
 0000000071 00000 n 
 0000000834 00000 n 
 0000000954 00000 n 
 0000001007 00000 n 
-0000009848 00000 n 
+0000009853 00000 n 
 0000001142 00000 n 
-0000009911 00000 n 
+0000009916 00000 n 
 0000001278 00000 n 
-0000009977 00000 n 
+0000009982 00000 n 
 0000001414 00000 n 
-0000010042 00000 n 
+0000010047 00000 n 
 0000001551 00000 n 
-0000010108 00000 n 
+0000010113 00000 n 
 0000001688 00000 n 
 0000003674 00000 n 
 0000003782 00000 n 
-0000005672 00000 n 
-0000005780 00000 n 
-0000006800 00000 n 
-0000006923 00000 n 
-0000006957 00000 n 
-0000007150 00000 n 
-0000010173 00000 n 
-0000007345 00000 n 
-0000007606 00000 n 
-0000008098 00000 n 
-0000008316 00000 n 
-0000008597 00000 n 
-0000008881 00000 n 
-0000008994 00000 n 
-0000009104 00000 n 
-0000009212 00000 n 
-0000009318 00000 n 
-0000009434 00000 n 
+0000005677 00000 n 
+0000005785 00000 n 
+0000006805 00000 n 
+0000006928 00000 n 
+0000006962 00000 n 
+0000007155 00000 n 
+0000010178 00000 n 
+0000007350 00000 n 
+0000007611 00000 n 
+0000008103 00000 n 
+0000008321 00000 n 
+0000008602 00000 n 
+0000008886 00000 n 
+0000008999 00000 n 
+0000009109 00000 n 
+0000009217 00000 n 
+0000009323 00000 n 
+0000009439 00000 n 
 trailer
 <<
 /Size 39
@@ -333,5 +333,5 @@ trailer
 /Info 4 0 R
 >>
 startxref
-10224
+10229
 %%EOF

data/site/build/site/en/neuralNetworks.html

@@ -5,7 +5,7 @@
 <meta content="Apache Forrest" name="Generator">
 <meta name="Forrest-version" content="0.8">
 <meta name="Forrest-skin-name" content="pelt">
-<title>Backpropagation neural networks in ruby :: ai4r</title>
+<title>OCR example using Backpropagation networks in ruby :: ai4r</title>
 <link type="text/css" href="skin/basic.css" rel="stylesheet">
 <link media="screen" type="text/css" href="skin/screen.css" rel="stylesheet">
 <link media="print" type="text/css" href="skin/print.css" rel="stylesheet">
@@ -115,13 +115,13 @@ document.write("Last Published: " + document.lastModified);
 <a href="http://forum.jadeferret.com/viewforum.php?f=3" title="Ai4r Forum">Forum</a>
 </div>
 <div class="menuitem">
-<a href="http://ai4r.rubyforge.org/rdoc/index.html" title="Ai4 RDoc docs">API doc (Rdoc)</a>
+<a href="svn.html" title="ai4r Subversion repository">Source Code repository</a>
 </div>
 <div class="menuitem">
-<a href="svn.html" title="ai4r Subversion repository">Source Code repository</a>
+<a href="http://wiki.jadeferret.com/Category:AI4R" title="ai4r wiki with How-to guides and more documentation">AI4R Wiki</a>
 </div>
 <div class="menuitem">
-<a href="wholesite.pdf">ai4r project PDF</a>
+<a href="http://ai4r.rubyforge.org/rdoc/index.html" title="Ai4 RDoc docs">API doc (Rdoc)</a>
 </div>
 </div>
 <div id="credit">
@@ -151,17 +151,20 @@ google_ad_height = 125;
 <a class="dida" href="neuralNetworks.pdf"><img alt="PDF -icon" src="skin/images/pdfdoc.gif" class="skin"><br>
         PDF</a>
 </div>
-<h1>Backpropagation neural networks in ruby :: ai4r</h1>
+<h1>OCR example using Backpropagation networks in ruby :: ai4r</h1>
 <div id="minitoc-area">
 <ul class="minitoc">
 <li>
-<a href="#nn-Introduction">Introduction to Neural Networks in Ruby</a>
+<a href="#nn-Introduction">Introduction to Neural Networks</a>
+</li>
+<li>
+<a href="#nn-example">Modeling the OCR problem using Neural Networks networks</a>
 </li>
 <li>
-<a href="#nn-HowTo">How to use Backpropagation Neural Networks in Ruby</a>
+<a href="#nn2-example">Implementing a basic OCR application using ruby and AI4R</a>
 </li>
 <li>
-<a href="#nn-example">OCR example using Backpropagation networks in ruby</a>
+<a href="#nn3-example">Results obtained with the AI4R OCR algorithm</a>
 </li>
 <li>
 <a href="#nn-custom">Customizing your neural network in ai4r</a>
@@ -173,7 +176,7 @@ google_ad_height = 125;
 </div>
     
 <a name="N1000D"></a><a name="nn-Introduction"></a>
-<h2 class="boxed">Introduction to Neural Networks in Ruby</h2>
+<h2 class="boxed">Introduction to Neural Networks</h2>
 <div class="section">
 <p>
         The utility of artificial neural network models lies in the fact 
@@ -194,31 +197,9 @@ google_ad_height = 125;
         Rumelhart, Geoffrey E. Hinton and Ronald J. Williams in 1986)
       </p>
 </div>
- 
-<a name="N1001A"></a><a name="nn-HowTo"></a>
-<h2 class="boxed">How to use Backpropagation Neural Networks in Ruby</h2>
-<div class="section">
-<pre class="code">
-
-# Create the network
-  net = Ai4r::NeuralNetwork::Backpropagation.new([4, 3, 2])  
-      # 4 inputs
-      # 1 hidden layer with 3 neurons
-      # 2 outputs
-  # Train the network
-  1..upto(100) do |i|
-    net.train(example[i], result[i])
-  end
-
-  # Use it: Evaluate data with the trained network
-  net.eval([12, 48, 12, 25])  # =&gt;  [0.89, 0.04]
-
-</pre>
-</div>
   
-  
-<a name="N10025"></a><a name="nn-example"></a>
-<h2 class="boxed">OCR example using Backpropagation networks in ruby</h2>
+<a name="N1001A"></a><a name="nn-example"></a>
+<h2 class="boxed">Modeling the OCR problem using Neural Networks networks</h2>
 <div class="section">
 <p>
       Let's imagine that we have to implement a program to identify simple patterns 
@@ -228,12 +209,14 @@ google_ad_height = 125;
       signal).
     </p>
 <p>
-      We can take an example of each pattern to be recognized, and train a 
-      neural network to identify them.
+      In order to solve this problem, we can take an example of each pattern 
+      to be recognized, and train a neural network to identify similar patterns.
+      In fact, one of the most popular uses of neural networks in business 
+      applications is OCR (opticar character recognition)
     </p>
 <p>
-      We create a network with the following architecture: 256 input neurons, 
-      128 neurons in a hidden layer, 3 output neurons. We feed this network 
+      In our porposed solotion, we create a network with the following 
+      architecture: 256 input neurons and 3 output neurons. We feed this network 
       with 16x16 matrices (in fact will convert them to vectors of length 256). 
       Each pixel is represented with a number from 0 (white pixel) to 10 
       (black pixel). The output of this network if a 3 vector of dimension 3, 
@@ -262,81 +245,56 @@ google_ad_height = 125;
 </tr>
     
 </table>
-<p>And we repeat the training 20 times.</p>
-<p>
-The results we got when we evaluate patterns with our trained network are:
-</p>
-<ul>
-	
-<li>Evaluating the training patterns with the trained network:
-	<ol>
-		
-<li> 
-<img alt="Triangule training example" src="/images/t.png"> 
-[0.98, 0.03, 0.01] =&gt; TRIANGLE </li>
-		
-<li> 
-<img alt="Square training example" src="/images/s.png"> 
-[0.00, 0.96, 0.03] =&gt; SQUARE </li>
-		
-<li> 
-<img alt="Cross training example" src="/images/c.png"> 
-[0.00, 0.00, 0.99] =&gt; CROSS </li>
-	
-</ol>
-	
-</li>
-	
-<li>Evaluating the patterns with random noise with the trained network:
-	<ol>
-		
-<li> 
-<img alt="Triangule pattern with random noise" src="/images/t_wn.png"> [0.98, 0.01, 0.01] =&gt; TRIANGLE </li>
-		
-<li> 
-<img alt="Square pattern with random noise" src="/images/s_wn.png"> [0.00, 0.96, 0.02] =&gt; SQUARE </li>
-		
-<li> 
-<img alt="Cross pattern with random noise" src="/images/c_wn.png"> [0.00, 0.00, 0.98] =&gt; CROSS </li>
-	
-</ol>
-	
-</li>
-	
-<li>Evaluating the patterns with line noise with the trained network:
-	<ol>
-		
-<li> 
-<img alt="Triangule pattern with line noise" src="/images/t_wbn.png"> [0.62, 0.00, 0.02] =&gt; TRIANGLE </li>
-		
-<li> 
-<img alt="Square pattern with line noise" src="/images/s_wbn.png"> [0.00, 0.75, 0.01] =&gt; SQUARE </li>
-		
-<li> 
-<img alt="Cross pattern with line noise" src="/images/c_wbn.png"> [0.00, 0.00, 0.98] =&gt; CROSS </li>
-	
-</ol>
-	
-</li>
+</div>
+  
+<a name="N10057"></a><a name="nn2-example"></a>
+<h2 class="boxed">Implementing a basic OCR application using ruby and AI4R</h2>
+<div class="section">
+<p>The code bellow shows the basic steps to use Backpropagation Neural 
+    Networks in AI4R:</p>
+<pre class="code">
 
-</ul>
-<p>These results are very satisfactory. The network could sucessfully identify the patterns despite the noise introduced to them. In fact, one of the most popular uses of neural networks in business 
-applications is OCR (opticar character recognition).</p>
-<p>This is the source code used to elaborate this example (You can find it inside the zip file):</p>
+  # Create the network with:
+  #   4 inputs
+  #   1 hidden layer with 3 neurons
+  #   2 outputs
+  net = Ai4r::NeuralNetwork::Backpropagation.new([4, 3, 2])  
+
+  # Train the network
+  100.times do |i|
+    net.train(example[i], result[i])
+  end
+
+  # Use it: Evaluate data with the trained network
+  net.eval([12, 48, 12, 25])  # =&gt;  [0.89, 0.04]
+
+</pre>
+<p>This is the source code used to elaborate this simple OCR application
+  (You can find it inside the AI4R zip file release):</p>
 <pre class="code">
 
 require "rubygems"
-require "ai4r/neural_network/backpropagation"
+require "ai4r"
 require File.dirname(__FILE__) + '/training_patterns'
 require File.dirname(__FILE__) + '/patterns_with_noise'
 require File.dirname(__FILE__) + '/patterns_with_base_noise'
 
-net = Ai4r::NeuralNetwork::Backpropagation.new([256, 128, 3])
+# Create a network with 256 inputs, and 3 outputs
+net = Ai4r::NeuralNetwork::Backpropagation.new([256, 3])
 
+# Load training data
 tr_input = TRIANGLE.flatten.collect { |input| input.to_f / 10}
 sq_input = SQUARE.flatten.collect { |input| input.to_f / 10}
 cr_input = CROSS.flatten.collect { |input| input.to_f / 10}
+# Train the network
+puts "Training the network, please wait."
+100.times do
+  net.train(tr_input, [1,0,0])
+  net.train(sq_input, [0,1,0])
+  net.train(cr_input, [0,0,1])
+end
 
+# Load test data with noise
 tr_with_noise = TRIANGLE_WITH_NOISE.flatten.collect { |input| input.to_f / 10}
 sq_with_noise = SQUARE_WITH_NOISE.flatten.collect { |input| input.to_f / 10}
 cr_with_noise = CROSS_WITH_NOISE.flatten.collect { |input| input.to_f / 10}
@@ -345,12 +303,7 @@ tr_with_base_noise = TRIANGLE_WITH_BASE_NOISE.flatten.collect { |input| input.to
 sq_with_base_noise = SQUARE_WITH_BASE_NOISE.flatten.collect { |input| input.to_f / 10}
 cr_with_base_noise = CROSS_WITH_BASE_NOISE.flatten.collect { |input| input.to_f / 10}
 
-puts "Training the network, please wait."
-20.times do
-  net.train(tr_input, [1,0,0])
-  net.train(sq_input, [0,1,0])
-  net.train(cr_input, [0,0,1])
-end
+# Print the evaluation results
 
 def result_label(result)
   if result[0] &gt; result[1] &amp;&amp; result[0] &gt; result[2]
@@ -377,45 +330,129 @@ puts "#{net.eval(cr_with_base_noise).inspect} =&gt; #{result_label(net.eval(cr_w
 
 </pre>
 </div>
+  
+<a name="N1006C"></a><a name="nn3-example"></a>
+<h2 class="boxed">Results obtained with the AI4R OCR algorithm</h2>
+<div class="section">
+<p>
+    The results we got when we evaluate patterns with our trained network are:
+    </p>
+<ul>
+      
+<li>Evaluating the training patterns with the trained network:
+      <ol>
+        
+<li> 
+<img alt="Triangule training example" src="/images/t.png"> 
+    [0.98, 0.03, 0.01] =&gt; TRIANGLE </li>
+        
+<li> 
+<img alt="Square training example" src="/images/s.png"> 
+    [0.00, 0.96, 0.03] =&gt; SQUARE </li>
+        
+<li> 
+<img alt="Cross training example" src="/images/c.png"> 
+    [0.00, 0.00, 0.99] =&gt; CROSS </li>
+      
+</ol>
+      
+</li>
+      
+<li>Evaluating the patterns with random noise with the trained network:
+      <ol>
+        
+<li> 
+<img alt="Triangule pattern with random noise" src="/images/t_wn.png"> [0.98, 0.01, 0.01] =&gt; TRIANGLE </li>
+        
+<li> 
+<img alt="Square pattern with random noise" src="/images/s_wn.png"> [0.00, 0.96, 0.02] =&gt; SQUARE </li>
+        
+<li> 
+<img alt="Cross pattern with random noise" src="/images/c_wn.png"> [0.00, 0.00, 0.98] =&gt; CROSS </li>
+      
+</ol>
+      
+</li>
+      
+<li>Evaluating the patterns with line noise with the trained network:
+      <ol>
+        
+<li> 
+<img alt="Triangule pattern with line noise" src="/images/t_wbn.png"> [0.62, 0.00, 0.02] =&gt; TRIANGLE </li>
+        
+<li> 
+<img alt="Square pattern with line noise" src="/images/s_wbn.png"> [0.00, 0.75, 0.01] =&gt; SQUARE </li>
+        
+<li> 
+<img alt="Cross pattern with line noise" src="/images/c_wbn.png"> [0.00, 0.00, 0.98] =&gt; CROSS </li>
+      
+</ol>
+      
+</li>
+    
+</ul>
+<p>These results are satisfactory. The network could sucessfully identify 
+    the patterns despite the noise introduced to them.</p>
+</div>
 
   
-<a name="N100C6"></a><a name="nn-custom"></a>
+<a name="N100CD"></a><a name="nn-custom"></a>
 <h2 class="boxed">Customizing your neural network in ai4r</h2>
 <div class="section">
 <p>Sometime for a given problem, you will have to "play around" with some parameters to 
 	get to a solution. This parameters are:</p>
 <p>
-<strong>threshold</strong>: A real number which we will call Threshold. Experiments have shown that best values for q are between 0.25 and 1. You can optionally pass this parameter to the initialization method of your network.</p>
-<p>
-<strong>lambda</strong>: The Learning Rate: a real number, usually between 0.05 and 0.25. You can optionally pass this parameter to the initialization method of your network.</p>
+    
+<strong>Learning Rate</strong>: a real number, usually between 0.05 
+    and 0.25. 
+  </p>
 <p>
-<strong>momentum</strong>: A momentum will avoid oscillations during learning, converging to a solution in less iterations.  You can optionally pass this parameter to the initialization method of your network.</p>
+    
+<strong>Momentum</strong>: A momentum will avoid oscillations during 
+    learning, converging to a solution in less iterations.  
+  </p>
 <p>
-<strong>transference function</strong>: By default, f(x) = 1/(1 + e^(-x)). 
-  This function is called "Sigmoid function". You can see it like a 
-  "smoothed" version of the "Heaviside step function". It will always provide a 
+<strong>Propagation function</strong>: By default, f(x) = 1/(1 + e^(-x)). 
+  This function is called 
+  <a class="external" href="http://en.wikipedia.org/wiki/Sigmoid_function">
+    Sigmoid function
+  </a>. You can see it like a "smoothed" version of the 
+  <a class="external" href="http://en.wikipedia.org/wiki/Heaviside_step_function">
+    Heaviside step function
+  </a>. It will always provide a 
   value between 0 and 1. </p>
 <p>
 <img alt="Sigmoid function" src="images/sigmoid.png"></p>
 <p>
-  Sometimes you will have  better results with f(x) = x. You can change the transference function from the default sigmoidal function to the linear one, overriding Ai4r::NeuralNetwork::Neuron.f and Ai4r::NeuralNetwork::Neuron.f_prime (derived function
-of f):</p>
+  Sometimes you will have  better results with f(x) = tanh(x), or even with 
+  f(x) = x. If you modify the propagation function, you have to supply the 
+  derivative function too (in terms of the propagation function result).
+  </p>
+<p>To customize these parameters in AI4R, you can user the "set_parameters"
+  method:</p>
 <pre class="code">
-
-	class Ai4r::NeuralNetwork::Neuron
-		def self.f(x)
-			x
-		end
-		def self.f_prime(x)
-			1
-		end
-	end
-
+  net.set_parameters( 
+    :momentum =&gt; 0.15, 
+    :learning_rate =&gt; 0.5,
+    :propagation_function =&gt; lambda { |x| Math.tanh(x) },
+    :derivative_propagation_function =&gt; lambda { |y| 1.0 - y**2 }
+    )
+</pre>
+<p>You can also use the attribute accesors:</p>
+<pre class="code">	
+  net.momentum = 0.15 
+  net.learning_rate = 0.5
+  net.propagation_function = lambda { |x| Math.tanh(x) }
+  net.derivative_propagation_function = lambda { |y| 1.0 - y**2 }
 </pre>
 </div>
-
   
-<a name="N100F0"></a><a name="nn-more"></a>
+<div class="note">
+<div class="label">Note</div>
+<div class="content">Remember to set the custom parameters BEFORE training the network</div>
+</div>
+  
+<a name="N1010B"></a><a name="nn-more"></a>
 <h2 class="boxed">More about Neural Networks and Backpropagation</h2>
 <div class="section">
 <ul>