From 24e9d18d54c256b5d875817d1ba9d46d35337815 Mon Sep 17 00:00:00 2001 From: mhjensen Date: Thu, 3 Oct 2019 12:34:09 +0200 Subject: [PATCH] Typo in neural net slides --- doc/pub/NeuralNet/html/._NeuralNet-bs000.html | 2 +- doc/pub/NeuralNet/html/._NeuralNet-bs033.html | 2 +- doc/pub/NeuralNet/html/NeuralNet-bs.html | 2 +- doc/pub/NeuralNet/html/NeuralNet-reveal.html | 4 ++-- .../NeuralNet/html/NeuralNet-solarized.html | 4 ++-- doc/pub/NeuralNet/html/NeuralNet.html | 4 ++-- doc/pub/NeuralNet/ipynb/NeuralNet.ipynb | 4 ++-- .../ipynb/ipynb-NeuralNet-src.tar.gz | Bin 88051 -> 88051 bytes doc/pub/NeuralNet/pdf/NeuralNet-minted.pdf | Bin 539817 -> 539835 bytes doc/src/NeuralNet/NeuralNet.do.txt | 2 +- 10 files changed, 12 insertions(+), 12 deletions(-) diff --git a/doc/pub/NeuralNet/html/._NeuralNet-bs000.html b/doc/pub/NeuralNet/html/._NeuralNet-bs000.html index a630850af..9ed465153 100644 --- a/doc/pub/NeuralNet/html/._NeuralNet-bs000.html +++ b/doc/pub/NeuralNet/html/._NeuralNet-bs000.html @@ -375,7 +375,7 @@ MathJax.Hub.Config({
[2] Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University

-

Sep 27, 2019

+

Oct 3, 2019


diff --git a/doc/pub/NeuralNet/html/._NeuralNet-bs033.html b/doc/pub/NeuralNet/html/._NeuralNet-bs033.html index 099688291..befe0650c 100644 --- a/doc/pub/NeuralNet/html/._NeuralNet-bs033.html +++ b/doc/pub/NeuralNet/html/._NeuralNet-bs033.html @@ -361,7 +361,7 @@ MathJax.Hub.Config({

As an example of the above, relevant for project 2 as well, let us consider a binary class. As discussed in our logistic regression lectures, we defined a cost function in terms of the parameters \( \beta \) as $$ -\mathcal{C}(\hat{\beta}) = - \sum_{i=1}^n \left(y_i\log{p(y_i \vert x_i,\hat{\beta})}+(i-y_i)\log{1-p(y_i \vert x_i,\hat{\beta})}\right), +\mathcal{C}(\hat{\beta}) = - \sum_{i=1}^n \left(y_i\log{p(y_i \vert x_i,\hat{\beta})}+(1-y_i)\log{1-p(y_i \vert x_i,\hat{\beta})}\right), $$ where we had defined the logistic (sigmoid) function diff --git a/doc/pub/NeuralNet/html/NeuralNet-bs.html b/doc/pub/NeuralNet/html/NeuralNet-bs.html index a630850af..9ed465153 100644 --- a/doc/pub/NeuralNet/html/NeuralNet-bs.html +++ b/doc/pub/NeuralNet/html/NeuralNet-bs.html @@ -375,7 +375,7 @@ MathJax.Hub.Config({

[2] Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University

-

Sep 27, 2019

+

Oct 3, 2019


diff --git a/doc/pub/NeuralNet/html/NeuralNet-reveal.html b/doc/pub/NeuralNet/html/NeuralNet-reveal.html index f02cf456b..acebd4f35 100644 --- a/doc/pub/NeuralNet/html/NeuralNet-reveal.html +++ b/doc/pub/NeuralNet/html/NeuralNet-reveal.html @@ -148,7 +148,7 @@ MathJax.Hub.Config({

[2] Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University

 
-

Sep 27, 2019

+

Oct 3, 2019


@@ -1327,7 +1327,7 @@ The back propagation equations need now only a small change, namely the definiti As an example of the above, relevant for project 2 as well, let us consider a binary class. As discussed in our logistic regression lectures, we defined a cost function in terms of the parameters \( \beta \) as

 
$$ -\mathcal{C}(\hat{\beta}) = - \sum_{i=1}^n \left(y_i\log{p(y_i \vert x_i,\hat{\beta})}+(i-y_i)\log{1-p(y_i \vert x_i,\hat{\beta})}\right), +\mathcal{C}(\hat{\beta}) = - \sum_{i=1}^n \left(y_i\log{p(y_i \vert x_i,\hat{\beta})}+(1-y_i)\log{1-p(y_i \vert x_i,\hat{\beta})}\right), $$

 
diff --git a/doc/pub/NeuralNet/html/NeuralNet-solarized.html b/doc/pub/NeuralNet/html/NeuralNet-solarized.html index bf649ab94..c2cc3ebea 100644 --- a/doc/pub/NeuralNet/html/NeuralNet-solarized.html +++ b/doc/pub/NeuralNet/html/NeuralNet-solarized.html @@ -261,7 +261,7 @@ MathJax.Hub.Config({

[2] Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University

-

Sep 27, 2019

+

Oct 3, 2019












@@ -1333,7 +1333,7 @@ The back propagation equations need now only a small change, namely the definiti

As an example of the above, relevant for project 2 as well, let us consider a binary class. As discussed in our logistic regression lectures, we defined a cost function in terms of the parameters \( \beta \) as $$ -\mathcal{C}(\hat{\beta}) = - \sum_{i=1}^n \left(y_i\log{p(y_i \vert x_i,\hat{\beta})}+(i-y_i)\log{1-p(y_i \vert x_i,\hat{\beta})}\right), +\mathcal{C}(\hat{\beta}) = - \sum_{i=1}^n \left(y_i\log{p(y_i \vert x_i,\hat{\beta})}+(1-y_i)\log{1-p(y_i \vert x_i,\hat{\beta})}\right), $$ where we had defined the logistic (sigmoid) function diff --git a/doc/pub/NeuralNet/html/NeuralNet.html b/doc/pub/NeuralNet/html/NeuralNet.html index 99e3c4d63..f7f414bdc 100644 --- a/doc/pub/NeuralNet/html/NeuralNet.html +++ b/doc/pub/NeuralNet/html/NeuralNet.html @@ -266,7 +266,7 @@ MathJax.Hub.Config({

[2] Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University

-

Sep 27, 2019

+

Oct 3, 2019












@@ -1338,7 +1338,7 @@ The back propagation equations need now only a small change, namely the definiti

As an example of the above, relevant for project 2 as well, let us consider a binary class. As discussed in our logistic regression lectures, we defined a cost function in terms of the parameters \( \beta \) as $$ -\mathcal{C}(\hat{\beta}) = - \sum_{i=1}^n \left(y_i\log{p(y_i \vert x_i,\hat{\beta})}+(i-y_i)\log{1-p(y_i \vert x_i,\hat{\beta})}\right), +\mathcal{C}(\hat{\beta}) = - \sum_{i=1}^n \left(y_i\log{p(y_i \vert x_i,\hat{\beta})}+(1-y_i)\log{1-p(y_i \vert x_i,\hat{\beta})}\right), $$ where we had defined the logistic (sigmoid) function diff --git a/doc/pub/NeuralNet/ipynb/NeuralNet.ipynb b/doc/pub/NeuralNet/ipynb/NeuralNet.ipynb index cb077d392..424e9f9a3 100644 --- a/doc/pub/NeuralNet/ipynb/NeuralNet.ipynb +++ b/doc/pub/NeuralNet/ipynb/NeuralNet.ipynb @@ -10,7 +10,7 @@ " \n", "**Morten Hjorth-Jensen**, Department of Physics, University of Oslo and Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University\n", "\n", - "Date: **Sep 27, 2019**\n", + "Date: **Oct 3, 2019**\n", "\n", "Copyright 1999-2019, Morten Hjorth-Jensen. Released under CC Attribution-NonCommercial 4.0 license\n", "\n", @@ -1465,7 +1465,7 @@ "metadata": {}, "source": [ "$$\n", - "\\mathcal{C}(\\hat{\\beta}) = - \\sum_{i=1}^n \\left(y_i\\log{p(y_i \\vert x_i,\\hat{\\beta})}+(i-y_i)\\log{1-p(y_i \\vert x_i,\\hat{\\beta})}\\right),\n", + "\\mathcal{C}(\\hat{\\beta}) = - \\sum_{i=1}^n \\left(y_i\\log{p(y_i \\vert x_i,\\hat{\\beta})}+(1-y_i)\\log{1-p(y_i \\vert x_i,\\hat{\\beta})}\\right),\n", "$$" ] }, diff --git a/doc/pub/NeuralNet/ipynb/ipynb-NeuralNet-src.tar.gz b/doc/pub/NeuralNet/ipynb/ipynb-NeuralNet-src.tar.gz index b857d2225d17e8611671ba86d94452f9ca45524b..baee605a4762995613d43ac8dea6b387fd56389f 100644 GIT binary patch delta 21 dcmeyop7ry3R(APr4u+<4Qyba0vNOI32LNLr2z3Ae delta 21 dcmeyop7ry3R(APr4u;-4y^ZW!*%{x20{~;p2!;Rv diff --git a/doc/pub/NeuralNet/pdf/NeuralNet-minted.pdf b/doc/pub/NeuralNet/pdf/NeuralNet-minted.pdf index e11fa2a22b4ed9164745c3d6f1dd0d95e1b6382c..02b4caf8ddc3af96acf938d0a6683cb6f2133ede 100644 GIT binary patch delta 6469 zcmai&Ra6v!o5dxE?h=UssR8Lmx`&}*2+5&4q(ow%?PGSZxlPhPP}AGIw5|@knK`PFB6*7JD(Z6IwX1I|f~;R@T!|pSYw5x4AK~#D zZM71sQET{{uh(OP_AP=tZtk|5V|bG1*cQ`u7bQmTceJc$as7VVFFmR~MV{@&Y39k6 z5TpM3(-GCBU+LUa5DiV%yOeOA~jU4ywBiFwU^N&&u^jb#r zP)0kLrMm1J?1>8*zj8{MyNTmTW2H3Zq!DTTuE}mvQU3MQ=d&TZOk$7jAMgW$p|=@C zQ>c5Gmf}Wf7DEKeno*j3B3|$F&GyNs-z9EF!Zp)shDvpn6bS)GBToa;OgPhCNu4=3 zI^6|j{fhO0^#>)p1Oy3BZ_&pWUgp~SIoZq6b_Tw4Zev0xU+#3y?lwY`nvjfj&lu(Z z({s(=P-xF+o-U1-m9bc@919;e4qxNVD;?wF$UlY_qxn^&$mNR*tv6E;JL^Fr7bJzh zRJ{zIe$`N(B}mG+PgRO^g{7F+Tr+Y8kby+mY|0}e2;d2>+~qvADLPftBim?OYw=#! zpnCroMy0HII-!(U9*;hKr1$=0PJ-`^8~QfirSrz1Ilj-BwzPvsx5Ljnzq}NTS{$4t z=^{c&zWsE6Eg1_qaHMPgxM-`x+BB}8C|(=j#>USM$q^~GX|W^}N_ea}F94zoJr(5z z4-SOPQ?CbzEez0C!`6sP{7DCVBNznUHEYJd{Mk3QKt~kpx5&)Z(*JG5Mv*q_VbaDh zZ{6(Gn&!-`L71lQp*LKzN1}0Ycd~GOTz;D=WY1e-G!;hn*N5Ou7DF$SB5s^ycJDMH zZ(fSck7NRyEje@Zt6-L?YXnj~?+rm$b4I#7J>d|;ehz6vsxCg-p!Rf;GT~s%Z$C2E z_qe>z`dE0s60Hclk=?*N|ImgeO-~!x-}P+KN-TP+q?1)&iJu?NjZdYMDvfta>u4X# z(C& zq-2J*#JmA^$y)Fn(W(%cJpaS?H_r0Urw&&JPvP@CYJPErZ9l^$KXbKtWql^wt|h5p zd`?!S2G0CF>|&@kOozBuCgzw9{}Y=w|AqNV?a9i%_cBOu-~qLd9m#4CaEM19|18*o z;^f{Q`_eVC;eL90UhO!0G3rk{M5^5M{BU)hmj ztGfLRJS!T0_l8fr6jFW5G=9z!&@Y#CDA6pRIOu3O<9Nl!a$ry=P1c$3CZpjQY>GLM znb$OEmp5>|94!f<2icACxQ?l;&FHDVhx-n!1Xwv`*7ql#YkNgY3+Z_+OPKNaSf};R zKr}I`Nd**0DSWA>FoX@BQ}A>V_?dgX?yAl=rHh1fc+x~j`H3O>xJf98diuoKf0<=% zj}}zyj~q#x>Xz{qp*Os3HpN~J^J}=9_Nl$G(7|$i8<0rmv|G<{+F2s4{d}0|w0Pyd zXevKjEL&%8;vW0MsQ^+O@iqEo8oxb$H;s=JPeGH>AcK}?R`9B+DG&A7mukml@*H<9 z%w#Y;T8__z8TIGa+glFmT4Izh?k}?Itw~EW6WdfU`}<-V*3rwfpS)2t;&C&W>EY_3 zVm^Od*Q|4oqj>7nR`)Nsa~b!&n;WueCT3%s`W_TY)+Kq#?#oL_rd%HQs)DZrc0E|;(dlK#abNMYqvOBF816)h+8xkHTbIg zM%euI(>LZ{ks&DM|iIl|XQh3aNS{PpB{WbnJTM^j$J&F=0 zmbz8rJ#b;Gl-$Q8X+iEo4j_Z(zD=%w18)g#L`vI~co6kg^*+hipD82tR@kl4&6*C) zpxOSK6#}E<=D0J?dqZFKeauZOzT+Rw#l7&t4rHCgiokUy`Gl-mk?k@MhLIgW{U;og z2T1)gHtjk&rg;R+KPq-=1U(k1Z0k1`y|@b}=so9!7_HyeDH7HS9pvzH(0=g{liQsk zIMMlOk||^}6-A5+?8>u?i9z+tgt*EOJ^*5cg*t1-k4$ZK_r>2!hP0$ue_NPCr2bgz z-#r@gdqT%AmMg0d!tqH*6H7~6Ld-GQla*nEmpi%SJH+2*vL;id1&TCsPS)0f1qGF~`Z~e~aniXxa zpM!0E{c=bn_z`|$FQ0GHg5(836#42le#^&E_osR7W_*L^^8={V!+ZOHAryS*{gctx zk4;Oz?~hPv;cd?McV@;__a9f+f-iU6pQZ0)Ts;>5NwYI?5qac_+RR(19OLa98fZH1 zHp$&7X@ft@3pE`~KuRQ;0?VZuD^M!q*Hwr6RyU7p7W4vm-cnflcg26ZPKn!!5Gu|1 zlm3qnm3IEmrOP2t#&;eNXfWL2-pq`u;))@E_3kJSH>ZJkQw5{&&H_lgA)z|-Hl!Yl zX&8s(3@@4|Y}|aEC*%PYY>@`9@vsPjghBwSl~o&pKbNvxEPwPcen)HIF*i@u#e9%Zkof8OKbceuTS`2vDq1QNd> z&g%Rw)qDFbLd zCLN1=CbTcfX|MWD0e%8^xjZU&tSIWL81pd;Bf3O$(c>`*!CVx!xMkPU6oUD?H?A%z z0sG%LjBX%oqHA(9C*ODALN}LE|2E;8oCaHdj#nZq#wF3<6QIusn~rwIxXz4wzzKuU z#G=tG^-#d*m6@Zl@@V4lk~Uy; zAC=M2py#o;g#CJn%Yz!{8A{XS3}= zgXPvz-!!d%bkh08KfJ+tMPcK#*QIxX5Ba%uMIHTbSKry+?Z8JqA&+p~da zYHPy+h%48p79sv`M-A^5-7!`sMEyj&R=is|`)FLWZB0S_U&d3HGZ=$=7xB#HA23)W zC$s7HMP?HwoSZK*j%7rjP{K}(w#}(OqDhu&;m=hXAm!G0&wKaAkW)JJaFMw)@$&6@ z{h__mPsAO6f#UsmRP(V~EtZ_!C(pN67;01Uxv+dIlKE(m@_(JzrV<~|904FJ9ZW#X z5MAN79pK(>OyH;Rovzk!IzsB?muw5M_aB2M)pMp|5ch&Zz3Xk|yoHPw>`hY2nv!Os z@UWy<@&gmyAv>lN0&h(eEQ5)y7@wTN%R|JM5@P`7;AlFwOM{@GBhstW%AX>1R$-4z zl7}=Mo$0qVc5&U?nw*ro{&n6jT&xbM1gw9>TX!wbQ5bc!s6MiSro6r@Jed|6JC!Rh z=#wK5q4uQs7D-c>zn-FX65M?4cXmAuKlXzYX!PQUq9zn=#%sdI2z-QFO@6oOVHhZv z)eaw_JoR3U>6Zm3=Z3CI8~x_2U_9Jax-v_=g^w!2d2 zxt1uRBj9(ON7xG@%1`M19gU*Y`XBa+_ z=_BBZI)UaB3uB-l8E5>vSmjJOalc3hpnrkkGdtb~RCUD-%H->$es|Y6GUF`atO3*3CdK<*|ZSk&z3tC3)kT3Wu?m zo~;@48SqK+$3t!IV_rO?p%Z{zrOHP;VuqEAM;N*wr;Z%5tN}L#5WGI!n<<4yyi1|o z=}GXS&ya+QexxLu`;m+d)0RQ<6~U^qzK0O>h?HV}angp%aet^hmkJHaLGdjNWN8WJ z_wdF#)EK{{bykG{z9k*vqdBxc*^fzDOD@ZBmi%<+GxFEp zve)-IF5rjHVg!Nh;w8IgJZyOAck(JY&CyJ#y%f$FKsmogZlG+rn48 zhu{x_(!A!&sSw#9^K&*4H$B`4F>F<+6RO z?*Z53XrD^+9w_!itUE+ny#S~y#)TWxFBonU=48Kb1Csfm@!T!rdoo?(_1U3^Ow5oP zMz;#~@Y5qwmMP$pN*Q`l0p~=2ygD! z{b0a$hhC#w&o$cMJ6MqDOeX+Ni9Ov1HYT5H`{MtO>aY81Y=9^_WAUn^q?Fz9vA^gs zLjnF1K8$*ST}m2eAd>e66lAkV*A0!tJ(VrMPwyDU2(I^>n>dq^6tJ6bsPCZT%+rie zya?%@d?GdpjB*IqsfyC6Li-2$KQR7*`46mrU^nYj0UpA*`T6(+`2_?8g@n2I1lalb z*g<%_8tyiVFmGERhoXc4AHVqjMw$coaeA=|Y#N_2f z6vX-Dh2;bkrGb+FZ^R6bSKZdl-rE8AuRSev2onTn{RvTDfl>^e1w0CR*{eYuA)sVG z|6~-Mnty{fhqf9%;=3*$cM;ZyK?FZftmn1{xi54!^VY!>{5|W1nL}*Mi-MRXn(f|! zIT{V2Z~OB3J}~vz(W{!9o$>qMPmfPtQ}F?wldVr=>~g?R{VmhMn6|t20dGvnMYl zB>R#sl1&VKC|7Bvw#%i{z#-h1?cdrGcnlAb^zPN!N53mu-BJ#2!hN4heY+vHV09H< z`G8f*)F1aue?w{qVKkQcok=WNZ-u79-ZP3~ECc+{!L% zV4~h6tL7)T{A1ig;N?NTJJ8`Ve`*v`;Xz2%tPNR3w)}l9vV*-6VQUl$ZNYM0*@VAa zidF6Eh(tzySHdZ%YK8| z*|M3Gg;iC}+}oh=m^{OlK{W}<#`^i+ty37>7fB*zh7M;gGD#$Jr#4VT^%*2`u^N73 zD~ntoNwL2}8LNgz$F4QuAAgo+t^0#W+F6XcT;&;#R)2yqdevayS+XA^t&YQ{zSB5iEI=Cfo4$^>c)(1@R(i-+(JS#COXU%_>BLganj zgEj3YA6%%1)L|EaUA9miSiuti*x2JbjX{reS*!nzn#NVh!~Y~F>S6FslW=NsBY^d$ z{-@h(v`p;mld&Ak0G6SMX1#3zN|jg{LYIu}zQK4^)Reslep?+QV3Gn(B<3Rdx-kbk z!&Xgq@#Hm6HvtLHRPis@6Zh`S!Vl{Wo`#kTRPYT>({QC0B1n57VJE2U(4z89B&89; zq=v2aJJWKxQJ!qNJhoA>WeE#Eegs>MQ|JyRlHBN6HV_>bA7L5AG=iH-XCO&ZRYN?i z4O9pvz|tjC)dY%%?nJx=H}WpyLsgyLMnO}MxBInFD00%z7|MMRL;STo@#1|Kv=JE$ zFB&OVK(4)8E|&nqrJbQ2g?5btz*jSqBE)f;(KM;(7ZMzQ#exdc*}mT0cLMB`{(7?((-EJ5RF zUg)2=tCUXtKcZ9Vmg>Ny&^_`0h{uJO8S-C~P1)V&?lSHeFT)v^(?TS1fVe{9c&w~S ITFQ9;1!(F(%>V!Z delta 6476 zcmaiyRZJWVgGG_S-QC>>cc-{JMGD1b(7|PpQi^+#0>z3u6fYFF;_eJExJz-^@B4R? z-M4)>$;r)4?#u0Xpe=u(O=5xW$Aya+!EkzCK@?|NYR_9T;>s zPZOsJE|2r^zq-#k=+w5gP3;v;|HkgW#sX1;Y^smjIjd!KooD|AVHTZayYj>)LdIRK zy#r8lNq+lwk+o0#dJee08*(EsXRXv|5d6{LUHmZeJ*tH0WI_x4lKy(zH%XBj9;QjN zd*gqQ*pQ8#{7jlqy1)-AWk01q)K&f-TC691^6W_d$h~n2>$rPmtP~pcn~ywLeoq7( z!0Ms;61u$~FX*_LS32eUrg)R=0D11605+&N8IY9FQka(JME(JK>xf!i*3&R+6&eb? z%u3gZx^&;JqvE1&ha?kLaDxVVSps^9 z>Kx3!avm7&i;9?VI;GD)VDV%y5$UswxOo1^)oeC2Frl&@{GM$kCI_!q3LzjI1FcAp z-8~_~<0*Aq%F;^vK>2%~fuEys+}xi1U7if!F{;4^jH{}2o^FB<-NP(!Xg6!c=SnfG z{YB*p3jE)fZ}hy-Fhe0C41bIn6eeFgzcJ3ZUqL~Oe|do(ryYA}Ux*4zvmT)!JyGl+ z+7L|fZ7)h~jotei%gOW1=6;#cyDvxto`*(?2p(8qX&BzD^N{`&M@s0RULOs>#FR zyx>g=$fxlk7U_SW?UFdd^`bD<(=k9w^T&q#&glBzWuLgOoz#^?u*LdLd9R zMz~cPRwH6Fm>6!Ur*r#bXpwb<{4+*AS2*H*(3uaRL9#joJVqHitV)B{60gsnK)vM7 z<-@uQsu`6~3>s5s?^2uSlU^gr1h6TmR7eK;I>~(+F^Rl~pwCWWx;5BVT_?Y$<|}Am zD-6We(`v4Ec6w@F@lN3fW-=`#gT2vDi#F=`#T3>c^ko+)JoH0KPqxmCuX2w#Y;LE9Sq=b!T z^h05yg|V4qz$QUGs1F7ABFz8uu@C*&ubFP+EcchGxXZ4fmApRuN<@RmWO2B2d&R)F zEJU~=5M(<9DI>KqW}B}`CiO(>ojMdQSI;KxW7q_QKcJj5dx$(YMSOtmQC9@4j>?%W zX(qFSC%QTL42!efEmXAqT7*93K@!yRt=2bLU??HqyrI43+}6j?cCl&?^$RR%R4V~q zKXfRM-}T%SjVhV3+ZThj@yWjrZ=oLxs&h(dT~_}CFBI5Vd~LBcru@?SRGh0WW}7ik)u)80S?NSYm32R}-yIXpH2h?T4!i zvaF_Gg%tCC#1Kl_?l-^~&bCGvbGyqvo(b{_4sz=3&dgC(;Y{eSHEU*gZ7eN&G>w!) z1rheN0iV(b?vx`9$Ro#U$8#6WCe9mKNg^S#s)Wc}zljnK+|dA^G%nmPRs1vd{VosK zovs2xI+b9s)pYH1Ym*OS`lhoJpjJBGrDOUlg*(_{4h^kZB6yZIL|N7JF%Lt1z2L;kRg9&Gqk;)XRs& zAzBFlZd_t%{+K?Av3m_dL)E{rT zh~*bhJ_;48ZXDcjCJhBe2mhy(Kxq z^A1t(b$ZDEAtlPF<2~ad9-hBS@N^8sRZ1aA4zhF1s|R}2lD?_$d`EP4%~z`*9rcsj zNdk7&=2nmG&nx_dn27gX9zqS!tS*Tb!965IZWi^458A%YUX*xsx6|OwL>0;SaHNXR zyG7^-zU<%}s4M{=tU8!-{>*ebyO?iKJ4UD#lHhYGaUoiLHdScC=tqYdPEJCWH*0i8 z(`GYJ&KoV4g2HsR-)VGzwh`)jhKydhz#QX^^sFvI_ZM*(d^qOZE#5yaBVIcPdB1u` z8h;*1th+#-#p9~#THpva(5gMO*T+^QO;9D-JgVh*)_&xaJT#E%KIlW|jr&3Rm00^914qKs&Wf4E(|x$zTt5oj0cOv`Oitqix=M#r+!|0 zxRC8S7uM&T*2%lSt3564dJ1j++)J_-l3S^cFR2u-MC)p+PSUng$IvX?{P1|W-VL`{ zE$i0^LG{i~%@1dXz$hHub_Q=dAV^Q;0{9ws8*3(#K z?ALOPZnxtAqL8JtwT7XK(5Wi3ZX6tg`val*04hU&%)9x&HLn8PHA3DvT}rB59S;=v zg2U~O|MtFiU^Z89(?q&@w5+2_c=qv+b@iA}ViucpS>EAkYF+a~=xN>P9K2TUJ zc<7i$**-lvL#AK_7+N6n>J5}HFWf=Y?X5Ntnr^Ui!2Rb!*_t9SB}$VqP_aq@uQus` zee~q^7Gb&8%@_0{j&ir05o#77W z3|pFlO`x4rKkb0o%4Ex{k6RXX+k-D|rK;pz`@|S0uir`(t+;k-ae0-{sdIzpj=k4B z%$4fopM#^-bcV+(Tz7gcc|Y;-HjtMy6%ZptDM7=qdcB%|gSIF)tFv9BMU0pID+h(g zg@Vo?*r5A{jdj8GE|brt&0WsUb3NSQ)yFVuIvt<&Qs{UQiN>0gF{s;@2+e~yo8PZo z#7cs_v}MCmPUCrP@w6mX>CDF_6UQL7BpTOTfG;&WTtI;|_P#SdI5vEYH0~YgfmKB- zjy+m`ctN=;*OUzxls^09JA-1LjOs@jX=o~B?mA9sB?jRO0!7G!*5@T81_azbf;oa; zh?FtF0es3oS!HY=>`?8q^D*-*+2Wr})&`thC6Z_{dHHGOuHf3@1EbTVWCivUWSr^4 z-X=~IqyM@95jevI@zj|tWiZ9r_n5}Eyl$2dC~40RQL^!J6uxf<~=YrlAI* zwS3DhJwXm}j_$|uUYEev#s46+EZ zJZTQ1%*z>_Up^WPZqTD#lfP6JSIOj}Al%_A(AJMVR%@cIK6V<0{gqa3nQ;BO$#27h zVQrp7*ht#!w!BKbaPC0PK?7hk;OTWc?%QMxO*E#qLi!k^gwxK3$=oqW_R@a8I4>=S zCa$o2{sYqrGeBrI$yd|cj9?7X4{IRIDANoA&PS{1vvBL?wkmjkJgHS0VVG}igMJ)D ziBM>O$4WAUuZcx%{RMMmuwpA zqVa0E<}-O7jv7gkprfGu3auJSv3;5S`X}~& zqTH;^>?Ovm2y+&Lh$!+vI!a5-Q_bCc^qWn(Tg~0o3m@~g(mO~Gs(mY7S^@K|`}r+d z06YS(BMi_LO4jpm$~Opsy-8f)>y#p@w;)sN4T4w17y2Sd?oJ7xDY>XmJuT2TNL>? ztvd_%&XvphK1m}egX4yLXGM6|9b9lxVOvKxvguie3551b2}5e&pNlvc@%S~SQ~&ih z{}LoP4^u47G+*X~&61-dlkjw=xP;D;<0>~e^=oeNq*O?;Q`L4<*|19ffB|K6A-mY% z4fVxF2dqV96V*E{UMO_FqH+b8d&IHjD9iw)k0z)oHr$^v_+5eX2o{t_(g-}I5fJ{H zizh%l7UF5Y8v$6?fw5P~6)2BQ3Q}|h7?q+L%Z0B}2YOQw-sqi#NdulgNe1?)z!7O8 z(ByoA#%GG7jjkK9iw46-?a}JDhDIE&VBgK?w=4cUYs>yyCt4#D-x`n#^3+QU>Rrfr zf=eeyzQT@plFRC-RtZ#R?JMmt`YPM>JG|%MV;Dr=L!1zchGsRI%lpy1r@E+RTko95`K(kgPXqk#yZp*d9c2(kJzX7`rP)v2!HnCCYk?Ndu zf%ab3vM|a=cqud23YQT#Vo<_(&c$fFVP(;L()OqH)#|&1N&>PQSSX|w_NBjAO|6U(9}C+8JiI7Y1R7=+s=&6gwjMlPF7`rIeS{E|@8+0YF@7ff$jyvvr`M zwOQesC^-UzC)-!-W2Khx;3Jj0wN4yT>5^)b!crIbGw44DS|7l)ZKJ_1rtZ`JvMJi3 zCp9bP?5ASzm%aLzeS3Z;I$0L6jRE9IfSQ7=sH0*u!ce{j}lKdPVb z4#~QR&s*D{159xy^`4BrfMmRP*^ZmJu){OHu{7LE5h~x6#v@eC^UTaoxCjy1;&Y@l zn+ppTT)!YIYr&ycA=&EemLwq2tp&O=(=IWI;b*%;SVVplwdxfXh}KJ(3I^jUufi=` z!uBfCf;F{Y4!{#XIat4SMNwikiMNbs;}Cz8dRqkLkkqz|{0yNS8a%i9x{1?yKJ{VX z;l^*4aWtfudXwqJR2pGZYglc_%1^m;sER`4+#ae2zWl(nhTKs#WWP4TY{=pLh4UD( zeN<$N%JD|l3)g?j&8Mqf^c7LLXFV^|iW^`k_X{-EpHR_5L9hi$Z=oQ5GRVMJ;Jo?3Sp=jmASjG95dCsDer5s2=%P6cLWr`;$GIK@;!n)S$ zF&Cl8Rl7QB>lLlzEudkpRxDpjtOY2yi+j|LBjlRCC~~c!Sd+kmrqh{*tG+5bMTN=i zXdw8x(tKPFdpJ9VgA4cW@UuBnfa(pHCP|MA9-$PViI#&e!sM=6EI`v*o3{e^@1Q!u z$zwBp#Xz@IL4PD^A@&KX(IIoxgqQ6XNNgCL*8JrOR)9t4th&&IBGxY%(#H)(fSBkt zlat8HEE*jSC7uZ=VvMv1F(e7v9--uc}%XHVE;QflAaJ#{Cz2sm)2diHO-w3 zs@wZRL53wvnayOD1IVOnW>27w_;2`04J{)@IyY*pDr{XbDUFuRe;O${X42&=6Rn17 zcA6!-|1?oD^QuAG!67>?8)RHbm263)96iMwztE4AwZ?cQ3meD;4YQIW5oLPN7m z1=EI)g5v7SVI`4#;2J410=aX|$w~^qb!44?vQv3od*S`hDInNXol~SlPwD3=yR}pl zYgkGJMk%?AryWuUQ4B!12p}MbC{M4ik^kOGCGYDJh~c;6)^gXgN}ZCYx}UgUjjl)k zedF)uLs6IKR>rp=$HqF2C$=s1OOnvO!A89&(K-`5ru6k+JNSPPp;s>NHx^!rIaMo8 zs3?4~zFLyi^g*zU!@kvxs;z9Trc}HvSBTdkAG?zXF&b?+n-XylD(elyjS-%tG1`Sq z0Svj4a%4OaVyV7p(OMOto&Xs|ds6yhPO!An{Gc7Or*q?np3B;*7 zN+4(f)zs}hHmR>;|2D#wF3QXZXcb{y}(wUU_Vb=J9I&Q5q^Gt MbS9>ET8ilZ1=|8nM*si- diff --git a/doc/src/NeuralNet/NeuralNet.do.txt b/doc/src/NeuralNet/NeuralNet.do.txt index 0267f7d80..aed484175 100644 --- a/doc/src/NeuralNet/NeuralNet.do.txt +++ b/doc/src/NeuralNet/NeuralNet.do.txt @@ -961,7 +961,7 @@ The back propagation equations need now only a small change, namely the definiti As an example of the above, relevant for project 2 as well, let us consider a binary class. As discussed in our logistic regression lectures, we defined a cost function in terms of the parameters $\beta$ as !bt \[ -\mathcal{C}(\hat{\beta}) = - \sum_{i=1}^n \left(y_i\log{p(y_i \vert x_i,\hat{\beta})}+(i-y_i)\log{1-p(y_i \vert x_i,\hat{\beta})}\right), +\mathcal{C}(\hat{\beta}) = - \sum_{i=1}^n \left(y_i\log{p(y_i \vert x_i,\hat{\beta})}+(1-y_i)\log{1-p(y_i \vert x_i,\hat{\beta})}\right), \] !et where we had defined the logistic (sigmoid) function