From 90969f2ab36ea316e606cacfc88a74f4314457c9 Mon Sep 17 00:00:00 2001
From: Victor Phan <me@victorphan.net>
Date: Wed, 12 Nov 2025 20:52:57 +0700
Subject: [PATCH] apply CNN

---
 __pycache__/new_import_ODC.cpython-310.pyc | Bin 0 -> 24364 bytes
 cnn_model.py                               | 274 +++++++++++++++++++++
 2 files changed, 274 insertions(+)
 create mode 100644 __pycache__/new_import_ODC.cpython-310.pyc
 create mode 100644 cnn_model.py

diff --git a/__pycache__/new_import_ODC.cpython-310.pyc b/__pycache__/new_import_ODC.cpython-310.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..1d30f299348c7354f2cd56ccfc831e0ae3acb42f
GIT binary patch
literal 24364
zcmb7s36vb?U1wKyb$9hKJw1m;qr-ApwrsT{jZR;(e8`e~U}>ya@>RBJdb(<6dQ{y#
z`l@@Rnbaf(IV6OLI0ul!hEdpr6>K28ENs?>Wgp4HTb5mRVcCV+103Ey&;oCh2MOSf
z?fw1!Ro&AwQh=R#{nc0Bef+=w{Z(sVAeX@3<JsYQ>W31EKjusSpMjSL@$*L!a}ti`
zB^sKoYf(IDCne3W4M%sAb15fzm^p@*Ze;9CT*|bwZgwT-q#W&Xw`SfhNX-Fz0ME2n
zYz*3iQa<CA8bkJw#7%FwF=CHMJnM}%HrN{)WA>Q5=e&)LaeEx`yuHcY<Qj9E7c})O
zpDSD3O_wz(Rq(bpZm@4?Y_qpDZnSS~ShgkU1Kv%I?e=zw7rmPsx7fExT=RC=w_-$t
z-ffNB?As)*e?%p3XQOPFC4C6-+wI#WKJ2}xvCH1o*lq8Q+nKQU$PD+!&wciOc|Twu
zl;<7xo$|b^akqVs)EV&(HSV?VmH4Q)uW{HuEb$HAeU1C=`{Q;Wupf|m584m9`{#~0
zW6nlr{8F^8o4kh_NA06ha<g}=aoj#G@h#rNjYsT9B)-*qwDFkzSmS%_?~(T#yvG|S
z>=P2-=ACRz+LIFRuYaTWMB|iwO42Qqf3N*siQj~FpR}Kpcz++;qdp}4X76<4jD4o@
zl>L;vYu?lLGnm0G-m{J8?B^tHhi5mQx1X1I|IGVW<W}#6#*6lgQsy@Adm9zIBJrIV
z&-dBiC-JgZZA{ry62Beuuh}(;_mAy87@K1|4cB%hPxGejX<5-3dq$qKjk;Zr^X3|7
z?Xwc!<#~;U-DouJro8X=T8(*oUgFW3oI`xV+4FM3R?c2$AD+Im-#LJ1+d1gmfoI3L
z)42=J1?O((9z4%Ghn#!yyx<&m?!$A@x!-vJ&n4$U=Lnwfa~^Vz;u$!{oa1;dI}baL
z;Q4;%QRgu{Kj3_i^EjR_Ii7RUnS5EZzu$SnIfaxTaNg@YiRVS<v~vc}mz}4ar}6xt
z^NjN>o<I0V!g<cIUrso-Tc7*T#iadV=XvJ^<or<f`?6C`!>KtA+WUy(I@3t~VW;WL
zI(4Lc)M+_$&RN7RIojz&dHyQ##bl`MwR2CMK3+NX+#|=II9=94^JG<>bz0|}OC_i7
z&wJIyiAME8MYYcR$Q`M*n$z`}j&dvhfl9SjbA2Xf8dd*ny*Yzu_PnaM-Ab)BQ`W<L
z+(e~Ooez_0W(oz0jh5f8wHl39vr<$3(w055t%kd2uC=gOZSHZaetmbd)pn;^t+RgF
z2n)xNa?G8scD%M94qvD$Rb8yKrsmvQyW-cE+_2a!Ug@;!9?Eo@zT2+Us<l}RX<oH1
zETY>nuT^y_{(f(NwQ|1Ro~^W|r%?<!CW_3cXd<RlZM*Gy!$o?*ZFU-!cB_J+cznO0
zD#}G+c8$rUA8}g`wJwCosn!K#=alPoYVCTf$rUkL^KJi%77j?aXti3Iaa(q3s&&?#
zeYmXIhS!>Dq2N%h>eV_Px~<fkP8|!1#IYx<O{dj(xP@W0k9t+#uTR%qg`84bRqM@4
z8|fA7PMrfOVjnuHT3f96HPi?<)Z!YoDu=^Cw-jeuI8mK)z2nVV%aO*+Q}ub*t2a5@
zfz$QIyyrgFnD4YDvv|6VK~$A<8Y7gH!Xv8goOY|KHhc8xvZl5Ig3x{HRIA=>hnZ6?
zZ*iv8L|e(DPoBo3c%tsNUFT@aYbgK#Qt}ViFSyQ`dhKk~V8c~2Zbi0~(;j9M%~qpc
z^&WPsZ9o}QOD8bGs(R9uO~QnbwxNH4;+&SkV1^rRwOR4cb*k8oimOyhg{Imctqk&W
z$FLPA01B=O2hO-nzonQGLBi0J+(P%nlTS2HaYw418YyY_Zyb?4en;{1U)2+dHYn>w
zP%J$4K>J+cH9dYy2KwhfPlF`V45q!JEseIdATg&qsX%kmmr!!q2n>H0>Sk8!K9g|F
z(}}=vvYQgiDdgpX)Le4T=(arry7_r+IbBaIXOLT9+d=wR;zfJe3`~?wEhN-W@)?+$
z63zfx$u8|_r&lsT_OkXdYdII>R!paOSzFEr`5+URhdKJez{Ge<L4Gca=MbJb^q6Mr
z9FdVQ5*L$NqG|3+_`!Lk^8Y#fo=KcLZY0|I#h=g;%LOg5R6=XRachO{?4Lnd$6PMF
zzi{q&^gOSfN6jGPUqShiZa=3vmQk5g#@T?e4+H~q*%$QX;__g-urd&E+$D((A~u9o
z7!EQk#mfn-zyXvqmq&t;xh!j))zu%6%FgJ)$h00{OotQae%;AB<KAGSv^<J+Ln*E^
zYw)=_vAiJ|of`^9f(<CQ=~8leEN)?JwT1TZ+z4753QErAS9NF0C4D)MQE%m_@w}lu
ziq+e2u^$??VdwL6V{6iGWbfTwMLkPqoqr~A8f)FozoIRkj8}9Fxi<w|OMxzo%=9)c
zZwxlBj9-===&mgxXLWVU+p8bY5{aCJKf=3p+HJORki6adtrPo?SVT#dQ&+%c-l8?F
zS`F*;0qf`!kN8%{2i~+)6^B>VTNBd6lP->9)3>Ttgn_O2)L6u&*1o;>+`adLgcc5M
z%L68EHKmWJy0md-wvGnofkA<q4@uH)8Jlk%IdQ_OE>!DYb;@(2v9aM+r)~N3z+zR;
znyz~sPje=&_>}MZ@dV<LTM<37c6Ntewx(MO1M=$CsXB1wVqCR58CipB+iIeR1sAhu
z^o>3)d$QWD&7!CFtZU83=<+sn3<~MDm0N9Cm=DRIteJbRgC%WXseP%xw4ZCwtiI)!
zZr7~%>WC*MCYFj8SJ%Z3G-upH*3yV&ov8v*wk;sF>LDyy(^^XJvQA*=VcOqc0ZjvH
z%U~d+K%qeb*4Y6p_4U_2YE8CeJ%QnTEJvs5$f95jaR=6<0-Pi|BM5Te_{mRx(Rxbe
z;@8|JdM~FyxLqeq0sFem(3tW%ZkTM=RhaZ^US~UNef?v<Ap6L9PJw8kqds`TltKfW
zID1w4$)E;c^6RgD&|1<lJ9C%iv=`^y`#PKAoX@WOpRat(I@;=Zjs@Cok%U?uQX9@^
zU3#^-81*R=#sog~GKw@8Tm@*KvD&S50E^M6gJ>$$7DJ76I|HiQsLqfErdkSg;fk&f
zqky^>0TR=-7LxtWO?b_5)YzM<1wM#fxQ^v_gbSMPc;4boZw+E<$FTfOpNs&eht0H}
zt|~yF=klFkaeB5o&yBs6h1nYCf5-y6F~5jC27}{UJL}Ee4Y$!!i{)>T@NnwKNzP<B
zI5lh`fT6r8OwHq1VB0;ATYF#{Y(#~5-0{$Cxa}(6!W7Py>xBgnNFS6{U=*Y=KILRs
zxV-={?6uDUhAG*XFnjz$&7J3g4)x}In63(3hAFUl4IgD|_4!40ABGqjORZKT)GsJI
z!)ajvp#herjY445{4mRbR2HfVQSdP}cdA-D8)jwr7574|JRa&!EzF^qTZwupd9CJ5
zyWZ|N-IO9IbN@S8ws@|->j9m)jy>8xNwRPRp}2bv$j%Z>E6pn8i6~0=1VpPEM$FaH
zT-ziE!52FX&I%K-(-)$rxd{H2Kw+B~;+T3G3lO)(I<g>k!6)xjnXcB_;6t)Nja4#P
zj(P^AuOyU%$6I$q4Df}!?%8wn^kYvx@leaf_MV!>r0&@Lblq)Zn0uP;`3e{`@NtzV
zjvbwtUko>LPBOj<PaD=N!q+dLP5)&CW}&E!YKETHbxqgIcZ^g~OKYY+teHszPgBcj
zMNQWWc$%7}ZGGE#sc2Z5`I7#w{x?S6Wa*@t%xO9NjAThGXv2CTIjWVkVZ5VNv~0YS
ze#bON(NhVnr1g?s#M4B}Xy-d>8UtCwL{V|UpD<D1K*Rwf<mk&Our^7uHXwJxGT{kU
zsY~|omJZV0x21c*N@;-+XhAwiIT^<US<GF^2vd}gNpYdA2bsC#8nQdkHjtKDlU9^8
zFgsv<1|^=6{O%}9LYgx%X*L(ngD{(5bx4?p$m+y|dDt28vdh^3N|PR89$7ZuZ#p9&
z%n~EpQy#6N_TNL8BN7~uXF2G#&XhaB9p72DgtB~B@@VZmKU=TOTIZ|2RjWn~?Im5K
z09EcI$CHq<4n_C}beU}fo7J1vQ)f<EJCC0_{m7vM`|mt>cbPb}y3ndS7T9{L?)x3r
zkD58Xo-|vaOgNr<aA;A`f=oC+k4Dje68^!0C|PfG8ZxD*4Ydz2@K$cMzbi?>&+0Cu
ztGf{_-F5<FBgR`#L;FV=6V5Zhy*<KNvaFMR*zS;ZMX!JR-FM%;N&pPA;8Us!Cus3Z
zIivQoy8{d!VP+1j2P7d%sFduwEKO+6s8(mbGPM|va(0y|1%a;F!?&FqP`O7W;x%1E
zA=D9+SR$E_q2kQWlPqi?S%{a@z0d{&amc!G-;%M*TD)&*$KORKNFk3w{4T49IC|39
z$#P!te1Rlj<<uFbKgIW?=Qg8B_9&b>$z0+i^(=yNPCduC&ER<k#3!;kVJ_m6xF)hl
zs>-Y>22!8cOi-F~5eu{ITvp~+@#-H!V3<HWKs`l4Jc2@in!5k#zcq72BtStxB*nFe
z#7A|gg;xJkh$it9L^6*E5Qqi@0x8SUHzigQfd&~tZ)*-v0C;l^^4Kam(HtNFXnxX8
z=H&$CvSSHIWXT{ABvBXg+huLVxSUuvP|p-OE|TpuCoAZ{fDR?s`8^CsPUBGzVy8g0
zHMIyR7j1}D^I9D+Lm)NQ6j+#f2)|yJy(T|-`sfL3=V|g;R#(W6M0T)ti;$4}4l9d_
zSPv#bdfaX8Jw)E_7>G@ru<wv{no={Ur?Uqt6VO7iwQ=Tt9(8rrIsyuBosyBnIR_4*
zcMOzl3E0|*aouelJY+rXg6Rac#cBTftDm$UdE$v<ah*HZce`5vM|@rWorkPvFbuUm
z{jNjSQ_Ze=;@WC=cZYJ)b?QJ9r&MbyX5G=!k6RBN=?>tYZplaMGqZn3xqVnA$Q@1V
zXkDp}zpj;iT(V~C{GP{Z?wYK!ag~z)RRY=7Gy}4&f{uhu<q?k-t-IHdYrMMaEN#SD
zeWcZjce}S|*3uqioCIK4HS7Qx$DIz|c9R8ion6A{E`r5}W{esbr@RK(p=^L$n1FPc
zNc<()!5IMUiza|wa}sBh3hIZC>*w^-i7T26QjXyjJ=7=mM$GKGvza?gW|4i}p9Kv6
zCEFiQEN$a#Dm4Q0#9^-m=Ec7sk#5nY&6ooL&H@h?akK*Kt^*9V4S7!rm9ges<%{Zv
znE!lBLEGa)jxg#FECewLw-lz%&$>`oW#Z!Hlpqv)cm~T?k;8(aK=%dCaj-9qV+>1V
zvMSI{x=PqMKTJm?E<!*uUBN=R)hYcMCh3nNFh`3?A|X>NC5Mwma6*QzUPeZZ>(Ko(
z5#gE(f_Vv%nZz>I8LEbxh$!ZODje;SsByaSInB|XB+^n)!NmDc`E=8CAeIzO9OsO-
z0kezYq=0+`^+*|g89Yv}jSM)IZe7HIaMF|5UZGJ8Ae$s1x?BS%M2dtHza(P@oEU_Y
z2tHPq2V&0qP$1Omt7(F`Re<CzfC`_hQA#Xrh3M^Jm9GmE<uur4*Mnx0rR>x+9}p5i
z_?wR6I4D6if!E4rXf&OLx}BqZ8fh;>9rQKDcRx(Fp_$J}U*N3sq8kK{&KhY+ZLAw;
zXi72$25nBXvm`GCkdRS?<M<UcLsPF{5Z9~{=O#N_K%|Y+1oV$p;t85m8<AGlF;G#q
zifoXQ`1%nH3XO}35h8EmKnOGdDNTJ6ku@O29Bz+vycLmlVg;QFonkD3STX?hnZ{98
zA4dVm^;sxkBBe_>Kq0Ni>6Z;VLlLI3Fx!nRKz#8RqHuQoU?YO;yUl&Ka+pMG;O((u
z>eGm>;nKM%Jyqwg;RUKTBLGpoOjX;nP_?C&(~bt-Ti=)<2fUnFx+~DYlc(qOxuk@K
zlXQ$rFfT$isehN8Y>@meIXO{Unn5<mO(!q2UDHXus)Ogx9H!ba4K-PI@q59G=fR8T
z+POgZeo-|d?fv<4X-A&qxdW2x^rW4hr>eT5$csSmhgKPSbTFJfZYWGcUY>0^Wj$hU
z9!9}(J~ScgLiRrID_C0-U;<s3EZS<Zzv4m_nh{@3vOqq9coz(lYiAxk10fNL1>v-~
z_2q*41dA1_&>=weAV98u97*cO5ZD4$vG6HUe6sFbfS50O_T*W2(a!Q%2yYc;7G}GV
z1&Kr)nZwp35HvUjN-nt-?tg{SFCzPQ5fHYar~&Z}!l1TU{)@Vy72hcuqdJ7RcZ{U^
zB8t=qx!pgaQ34Rt{NF;*PFzd~TGZxrfJ$;D`Ju$L<^UnSYz79oE5J5{yyY~08T<eg
z<gfrv)bM7ZnNF{yf}Eho8`|l1CV)tGk?3|Dy6SY0hdw*e(G;}CC>OnhS7Z6WL`sS!
zr1tK%NzFFVD>UuK3&wI`c|df?W-t&8LR{>omzIZ?hoLhr1SM#6i_q(iOs4@V=HxZ_
zf=73RljY2X`lgq)4};0vlvs~h0X)R4iokP)CLk(aXabvu5VN``f@hC%1Bi=86HB*<
zL8r2cO|D0S1ukbk`x*ZI!2?%sM^#{|Y-A9kIjjK132a9dADo-}C&<R%A3adcDsHo#
zeY6hQg1Qy;6L=4yE^4#KrVmwVL!CmJ`Uo>XjG=v0=a{esL3vbQR{dipyokV#=jBG`
zq%bYh^z9<lSkTkfK%6mcKh&PJQx_`GU)X7HiqB9x8)^&c!)PPSJR42dP9x$r=EKZl
zHvthQq@IPs2n2=sJ~T;-(JOPIe-!}#X%0!CytIA=VSZ2e7V@de7za+~KNT6G%pk<S
zG!8;h($rT_XblP^`XkdPxcak*K-`bWJK&qrL|ea@n5I3)nAT@#&!Gw<{do;x=JSvm
zoa~J5<X%ler2(=Zq@BD|0GS`Slv)NkUjaG|^vm!7ph9BVY^R-KU|y7WSZjD%vVh}c
zkiDoc0~ta!0aWOeF6z$EC{V65{F=x$WDKs1OsXpw;Z?}C?*jUM2JtZ6tjx8hz!!-Q
zO4+#{+pWr}t6aD`Pk;Td82mCou>Pkw%wWP6*b<o^HOJT>0|LEm%+#CpFhlrh%`cXd
z_V{X(l}fXM4S^;Vnl9h&nIOTQ0rD?Y9B2?{*Cy9!C<bk7st#HY&8`)*oDB8af}No;
zlLzJv<SvySLhIHk=pKQ1!A>4p${mTVo`<ZfJV5GKSWdTgT*;};e7}*wO$@dp2$O(x
zyD)`jD)Vf{uS7C3)aDR2YHcxR0_>?Ghy47{Sn;1T_yq>P#3pi6(8*OAz~vsS#<4LV
z%t>>V85l>yJVRhs+DOBEUuPk7i0nxH8e0>vl|aCDAUg75mR8en%t-hga>3BSrlz&@
z+vyBKJcly{EgdQ3lAFn4{5FcM;V_7gdr(WV=1>NPvTQ{Mv}yhTI1Z6+ArTwEnwsRL
z10B4FKw~@OM0O_dkmclLn3w5BhjJ-(tkraH0wbrle02|657X`iaQfik8fT%c3ex~N
zuRaAX1iI|dm;=R<RR*wRg;fE^<|@*Qu!!;%9~3ZJ(DiGQTg_VFg5*%gz`@OFV+hsn
z<F%GT(ES@igzs`({|XV17#L8=c{sq)LP++3zt-n;97X+-HkZUPglQEiFtG+{kQeeR
z;9fwQjO(Un+vzzRT^L%k(@7yBIGRCf#SC(DFt38I&w@un3Jx#Kt1EeF8>$8v8TGjW
zMuOO7eYwycm@5Va>MdbDq>*!Qr6fmwK*}YTi_xfH_GC#IK0&s!FgNC$JdjEOrpZ!J
zl3M!Q;M@@U!#g%?U{ZJ+80YWB5~$B006viT=a_(%q}*UY4d-_=2qjE|Ea9sMSW3!E
zvaTLvLdYV+>Pw72j=+W}=Xydxgd>V-ko8eNiGWZ66_yifXVh=-Jp(?ACWVM!qEKLG
zDMucmtpoO@ddu$mjX)kqLcvrX0!Z<o*`nVUozq9zl)f;F-vS)bC;)cPE{Tb43R46r
zLj5DWsxL4tQ-$clr2oLTJd7h%FQQELQ1lXeln~sg&X3qQ8!G-NCf0~*<fE_-MrCgA
zCve`1#tkUNi=ZXr5I%|k)2&(&s&%T_)BI`acS=c|Sgyw!PKube2e`z&A``@t08T9c
zb<V&BC)+82D*%w-JZH{94@JeflY|H<{M9l9${;lZAbVAxfg&9sl$--sB~OEzE`SXw
z5A6Pqb~e>6I3~C<u<jyUW+OY`8pem%#K4a`=C7(9z^?(TK=}?i<r?NE88JTup9n10
zB0J%FK8r<D%P6t5qXs1_Fh8s*!1Wb>cK*6d&n+kmQOKa7jNG*p2Y=(kl?IRtdIXzf
zwOJkrlQZ+M^w5k|A+E90u>MeC`9C?HKV)!>!D|f20of*D6%eQd)I+1@)#ud*kQ1h+
zA>Bc2wFla>ooaKo1$cv*NUgB$AG1Vq#+kN@^O56A*CsTw<{bcsVcNM^WT+!Ppxhad
zJ^1BNC%+D;;1INaMShGRHirIauOV>h2H2Q)jCaS<u<vN<FOao{bs@E;1Sh!g2T_|y
z9%@mrFTgi&JE4pQ9R;u)p6PaKB`w$o_|O4M?HKq0QvnYq9pc6Adlph)4y=7f*!CX-
z803M=OgOOQyH*fj-GO!%(iqgMf^**oT$T!oXl)Rn$Y%+Re;y3C3D6r_dJgOlT{{K=
zxCBPJRutGH99ce&I2}2>V0Z=0d3zAdkajlNE&=1AWb+{MP2_*@`uS0+WA@e9(knT%
zrlmo)84S-Q;YyhJpxz%&ClA9_WMw!QCWeRPc^@Rtk;Sh-@`Q`)QaYA6N8Uej?mkGI
zc#6cCTY)1;Fi;002fA>3kOOwkPA+A|&~eZHz4rh|6C7>*%zGnQ6Z@2~lcy#pm*|m3
zmjY`BN%%7#EN@VMiio1uULWkHfn{s69rY<D|0n}t;JNP+vL^cuD21#l@c+M=LD*K*
zg^8@H{)DlQFra)IraSY%tZsSK&OClh;N2d2+-<+NQ-wvcd7@RV1998t(I<`_ubez`
zDojJV#Z&z`D`#TR7tzPg!%_lR7VEm5p$J@QpR-Mg%r~3rB0K#l267N{k!v5djgfGm
z&$mL|n~JM?Q=xuNU01k8+YK;b?oace+>Zt$6J!p2eM$KJqJ%U{79iu2_cwv3O??c<
zX<XkX_6VfG@L7W7tG<OM)^JsnbTjzLf$|Xf?pzfZ07H()b{vl+Bn2EZvJG<ds1`I=
zvf>>EJrB5e<2B%R@M7?@UXv=;iE~2^xv=!)(%>0lphsKPJ<(oV(d=}yMH2Xyw&yKl
zqFtYUD@77!UCFEegaHb7rhbR9-(?_hv6C^O(z59U2^ivAu|g?g)Y8$md_(et2NRt4
zWmf&C3@$NPi_^or6pZlMZ*mh#C}fCbnf723MdNMtHWIHPeLXz&0&)XAS~O}`p<PPS
z!b=hdo=1dv5;1T=k+nAm{tj#{exg;CWds2f^9&r|Kmb#7X%M$`kfA!QoS(GEq4N+Y
zsC{79oyttL(WvgN!0$(WowItA0nePOBe<FXX%>d)j5>>Vn8x-qZ<Vqg;Oip{2otJ-
z04}ehLZ9t4&-$vx<R$_b80Q)1Y{Sefs1qd{ikM+0_K}Tx=Qzq{pAGJRvD)(txIlK~
zLO~tQs=jFbZWMclH&OQdH*N?CD0PxOTbsm83De&$4G4#V6QLl0g2~!<at-|JsGY=5
z;Gfh*)ZakZm$Vh_a$-flOgPmi)d!IW`3<fUk*7mw&>vI1#N<e&W*0>J3!nnkZLS}B
zpEqo*8~Y|l&;Ca>l}3CG0C71fpUSv4Bex4cIG@xA#Q^lGv7BBSj`*q|9dSI+B_-6h
zHh@d)fW|Tm{LQUE?a9S^f!b-{FCiCxlsKj7NY-|w)N>#aH0w80(Ua&NCBa3l5<$m=
zB@(Ux;I!bF0LdS+KLB^&)zoqh%1X-q=^*z|0<t}P{(`)C{UsMa8uxq)+JiQr+aBHq
z1*{YB>x^f0P1_}B=u(cYb;rk6S4UVLo)lAi*|;7Qcc`oS`g+GJhoW`^PJ>+!IXD39
ziroWO4ffVfdJNc*46TEAPnP8Xb+)ax7kMpp+XCFvj{pXv;cJ^(8b1l=aN5A&9p2aJ
zAxq7oK{=Lk7S%g^eE<Oz7HdzSoHz9@^B+R6bmtL3S_A$|beR%=N6V>+D~0$j(#0@x
z1xL4v?vP;3;7oAlA9URNI^$U{r{#pJEPE$VMmpajV?6&*{H|Kp*PiUi#GX3WaiLaL
ztT4rZbF=eupd+nC#4D`BziUtY{%`Tju!x*^asSb^^mlZVbdPCp^by)f?QYR9lRBMw
z@DD8Te|3{F$}7f5@tdTit5~CpJ%~3--!=-U|4l`2mNimC20a`25o7oP0-Q$Z8ED_2
zx)idE4%Zy8ptRG^!AwB$9D1-dDL<4D{V`2srUMUWTIWGarG+4gb7N4CE4KS6mw8}X
zp*sUbT`@01(;P&)T(OAS`njse#%qR3#6sa~ghO2waH!dZ?f@L@8tpD*jU<>8^&yl7
z1<p=;)fpeEt?3AjxkwiAW<Q#Z#G9@`73THp$u&5PUt^^<18TuU`F70+xtJ8e)CC5M
z47j^$sTY&iUjKJ@`-j$-v$Hokt`BwF6ih|2;kF`US6Fs>g|upsEKenh*k&Up@g=;1
za#h)cOg8qIV_i(!ui2$OyKL0r_p^x~U~rLv3@nFUA_Z;N8jJPVg+cnmC<9Xi*sv1(
z(uyf~mXQI2C;3{CBBCc)38McZB0&ObeE^cMSR&cL?h32a8pUiXR?NcWa9zc028vm5
z2YrfJIMG8f3(n&*6|p9`2<Uu(ilKxRGed?-G^k^tAqsLt-mzjfe;vgv+_BZa#8B10
zK~OG$(1TrrIp=z+MfGo4@HZL!I|jeS;NLT#-r~AS!$EQM6cwW=O77!&-h$Gwdg_0#
z;;_gS<^k>gtzmJAY-6P%<QE?f?%=URlKHOet;W@l*;{(j>B8P7d+e<^-}F)=_I6O%
z+tj4`Hu_V4&ERhs{4ImG5CESF6n%#Y-(c`%2H%wr9qi*nX=<o&O^AI9I)s5$P84?*
z*E0BoFCuM$uRlj5QuW7Nr}!7XYA)-LZj<;K)GhX`p}WDelF!V{X^@y<{z8m;MTRL7
zf7n3NU<M4JCY{$Z$f*XYl?<Q=@~pf;rVg@D@Jl`<Q$QJ{#tncnT=F=K(s&Nw3BLoU
zco=?)fWh2cex*Qzng2pN2MJnCXBbs6z~v36^qREMre+cLQyL8btwXq+k#|NuXE>uV
zOGCzO!=%DO!4+U%sO=B61EF>>)b0qiJ45ZRvbmHw(W%0dqY)+%v6Dw1okVbA5|Zm_
z+c*vX+@%6!LikxPT27}qV;fH*6y2lxESohb@xikfj>T}Lubo}m(357)Kpo%J<1gj+
zL=waVv_PWG7oYU6ef-TgK4aBpzxt)-taah5murWtw~`b4r{B_cKqdJqI<SosPn|d#
zDI!I$a~O${?&&8O6XBL_E@2YZ-|LA-VFpr=sHP`bA%(J$G*p1f2}2ZZkC?uooSZ!i
zcXq1sH?i1dR`@&vvQVPl6WPligt`;1U2WQT+X`e|P+<>IhXY-Ehgh<jgw#9HLYEYG
zNo^XQxFFad*J+q$={N_$RsJJwILx*XW7T$yrH1Nj{*3s|za0@bVZ&F*)*TJkf3red
zVFrCUYm<<gxGb3Z5U93Il9>xrM<*xu9qUlm6OXTb$1J2+f@rW8Eo*-EtDlFjXY0+^
zFV`*q%{RVeId8u41+ZAyuGhcNmP(H`A^uyU_&sFpoZ>|a@t@vh?b~Jf?zsxQ5og-7
zWhwlG_*<?k#|wC|3s6+*>dhFD`d18o7Qq^=7iUW>mc-Bh5v)fWPJi(0&?%lk#KxJ7
zPc@8pmvFKwsdjP&wg7k+Br2c<$QOSbl#24rzWglmc^Nu=iKX+%DL^}x8O3SF`#@aV
zgbkrvyC}6y$sd$_6O!v{eyJ~iXie+G>slX)YiIjvk4kM=9nkv*$;Tz8zWlMi{Edi@
zL%YVZo7T0x8Sh(AGkX~(pvu<1S~o-`rPj8%6m)hdb)%GuW@h!3x=BjGurH;y<L&0l
zpwyLH1kdLp+r-reu|HS2N$_9E<1S+jdL`UT^6dh6@!ngQhu0XAqdT6N`Klu#^Ve8U
zQ@8Qx5dyBh!~olyDDS{EgANXzx*2bwS;2iu*tRRUw9vHEyb`jH2#8NR{}6$35*BZy
zC8wJEqKy5F@3*rjs~1k;&L_*y^NH&W&@f|kp-ELUB5{QStZ*Qze30>jQH}k*g8QS)
zJD7P#6u*=4JMEz(P8Cwgf_r3PChl!;#)H43SE(+{RM>H^l>>G<x&#R;W^Etq-_^@Q
zztc4t`Q5!VbT?f)07UGbc#3yAybf;C8Zg;Rhq*dWWwfgKzViixN;EZSgWB#DT;#1p
zN3T-p<8Em&{&W{rfo)6D*{T4fRJv}QcJ?C5pahgb42$|wQE<ydcYm%?H1zm93!Rb5
z;s?m3Dde*J)5TX(K@G{R^`*j~3I2@vkw&{)GJ_P*=Dw0_HL}+Ar{;V03%wXi4&Yb3
zrv4yO#cs!TrPrajmIi&Xwlcyq2mAta2}+KxX>kKm#@4jBaZUa>@;9x?-`vaJ632Lv
zI~kd6)ayJLTh$|2Gqs1o1cQAH_A=Pd-~fX=7#w797lS()+|2;rGXU9SaEQTu4Dc<3
zgrXNkXe`v>yJ7GmjCz2nM;JWF;2{Rb7#wAQuNQzCB7n;0QO1QH3+0o*#Zbcx{7uf|
zcJaR^G~J4nF_1C3qm3o|pTm-ZR4{Nkhz3go5%k2h%SnYVK=e_Ba)u{gOwE-r2j2r+
ziNg(?N?52=a4Fb}b262R)2acq3DD{Z1mQpo?==?|!x-jexR0|4$1I5D6C5uTgHR4a
za7mSRN@|USr~6kzqzA-YKzfc)!lk-WeEg!A0F9(ksGuSfOfzChi9w2`p9#`vP6nxh
ze-$KJ`i<ZnYE4@25i=#m7NBhc*k`C~hS4S|F?vYR>SrRy?By^P<$wi6PPUiBSd;_r
z6C98X+J(3Sqc%sDmoy-LnthPQF%C!?2+?v;(m2{d%(#1f20mSLJ<S->(?Ma)dnPEZ
zz6Uw@VijpP31BXorqiq0c^T^|Ji^i0&YQ1aX<AcnzVXB4H-6so>TkaBp^k7Qrxwq&
z;B_Gdcf&g!SI#}aYJa&nkF)@H_voA)SF&J7!A<b4>w9!RR6Lzw@POxDl+oFRf#8PZ
zN34IyKzPv*dym6}K<-J(y^y|9f9>O6{V5E6bxwzthAqpAU81^-(^5&&V2$pj3`M2r
z5fgJ)@r94b`!FSjlyJl8vvA~A&9(RV)jz>-<oaFg^7TJB#>bhMg6mFg*3Lf+byNF{
ztcab%Wz(4`UM}@XuD{4yzrf&^82k$czl^{x#>2;rdt7~U?1K1jb>9XeH?Hn$uW+>f
zb}{z;0!z&O=vMB&iA}zn*KI1!9Ed!^;@tJj_Ba@JT8Ju?!7D%>mH9=P?H}?uQ>|}i
zc+0K(FHSai+YC3G&6?y00)n@!PNNd_6bfHKi80Twpn<Cvf<O7rm2dve$*EHhKu(zN
zoT*W+?Ec+@NKeN6v_Z5$brxhks;N^K(=TMYj*83Tm?(!IQXLLp5{CxdEBr|4yQ1Uh
zGkt_UvQTu5w&4yCjI7|^SpXgolK6&p9+%i7n+dG5@H0s-kF^UpT!YYBY-|q%8{r_y
zJ9Ek9aik9fa4cKig!D~F&rAAdq!)uBq;Em`mR|Z+NWVAWw++7=@w3`Reu!ZCCTC!I
zyHi}gxjh)EtMM#R7L7lPFzCSM3M4oP859V?84=a`=xd@2XqSQv?ZvNXFW$X;3+6IJ
zp5587yaTP?ir;O)__4%`#~^F(TrLMYqZJ6s(Nm97H>?b^7B2Ab#CUEGMuXdfGRFU&
z;5{oNlCxoDRGvGKvf*;#uEg>VtnyfJ``kun>@uxSyOwtcx1rr}XW+7i{M&+CgWW;-
ztiG}dhK{Jtt>-=y>}YQe#xH6&pao}RaO-FyxNS>fc>*Kf5|4ZjM!grmeduB9<;2H8
z8-vZ5!`*-99IhSvp59pZTxYEN`o_91*n7=b_n!N|_=Ll_0om?i5}$C$?!y}U5A=V;
zA=rbtY?is~!CdxYF8eT-{lN_@Hv~IGKd}w5JrcVSu?dM;i0zly7<ii~cTaHBS+uXt
z2YVuTz<O*0gdS)b!FFd1>$yF+#Tna}a5jxWqj6S8*`EvWxdZicfQ!vh9$;gOK*-h@
zHU=Gx!`TL{#f`y)W8ryIFK2tOM{;fs_V?!m+k>0WrqJ$J(e5qXb|Xl{m~IRXaO7AI
z?c7KD_Us_`>kj<xjQ0#oO|TVvcF+Iko*nG%*}?1V*`0lRc4u(MHG6i)E84l=yJpvP
zfAgAMJBVGwXD0xG2eE5+VAt-%uHA+C?ugIXt-)RKnnYLN2A$i2gIuk69c~D=1vdtk
zvl9lgyK!pn4u+?bAi3qvUf?RBmXCrueF1`gXAy)Jx2SqbfQ2{T_*}h9Kc%Sp8Y+S1
zdyHriU#X%~neO2`0k9y6tR@=nnM;51iz|Qq>tDW~7lNoy2-DH^XxoHM5YEXp`0SuY
z1vN7u(^D<qJp<Pt_+EDP2Vusa?MzR5?!#2*Oh;7qSJ8Qg3KV<-;Il7T554*N&$mSx
z@zWiPBzD~-)(oF_NAsYm$J_;cfCr|jLA;!#_Ze=PQ4u1!$LCwMS>Dww?y^LFw8#q|
zvO+C4gMQ^th`s2pCdGSpDI>M4=S902YA^D80B3ls1RqTB*;sPo;PleKt~e38Fxa}b
za0Ry_SSs-20baLkewVySp^k+PbFs<pcvaye9)7zeF6I`LaV`gIb*L`kM;$(FJ^tqF
zzuLCWVI_XbYOAmQOw($=`TCdUEo$HniK30{3E!-)LXK0#LP%OPT1b)cofr?J4^1#7
zxLS``G4f#Ts<grabGmP3D<YN8Vn$)wkwq4+Ibu6>5Mq3OEn%0ADztSxaxZ3w_#gu$
z<c4iPRBpt6w&MM(2%l;Xg9!$I#^5M}FERK80y_&ir3EuTF8-UmWJh5G;tTHE;vyO>
zf`H5F1+i_YNe(ROcN*gE9vUu2j~g{n?Z_QIC+_qWItqu;rJvP=@;!eKW5mN?r)5ig
zHN}?!;mqjqMrm7}MVhKIBim)4+BSUSxr%VMiJ2yRwxGu!*>IGPbGwX>@L%eWWbQrR
zYC9`SNTJQn`cS`898&OTlbcPkMFKW?d8L^VnRzzc&C2m5&L~h9+RMqyTQ4haf!Foy
zxRS)<FIL6CK>xQQo9ufZ5F<v^kMG;1E*^kGU8jlFu?zGd@1}++Sx7Y_>lh6@w%pqJ
z=(XKya|55P2+rVANhHiba|7jv&yQ~by0nUoZ%>HxH?+BR?!hb@mHBs@%yL}PpqgW5
zY1QHqU9pJoUa&>1BNwP!H79S;U<M@KJrQsLftM4$?XjLu9g64EY<3yc0cgtn|AI}W
z^>+B(TxuMjWS~@RXz6!z20Tq+*5J=bQca8taPo|<H;zH$VnS->e*yA!^w9_LgQB^O
zuGI8Vjceu*M;WwTh;#M7+4gQB52x!cYn;qU*&JFO7uJ}}aWoqJZ8IJ9w7#8jt=kSi
zK2g%TD(>`mavAkLENTtE-XArCDg1bxed;A>h>8`rxK0K_Oy5&@X25m~X<8f<A@PF6
z1$)z%jxRdF__8)VDW<*wes?i0X*ghBaVKLnCFOu^U6Yb_hVZWS>SP>z>cJU}>(CpQ
z<G}Z-@U^Jb{<`TKkv_gIeeF0{(r`BQje{vEXLH{;n9@HEzW0p-20B>0!CXVBCG6)G
zknyb%Jp>B3o__Q7Um-Po<CFN15vAvKB(eSG8$Y|2m{V%)QgqT*A>N?qxS%;BIw0<u
z9Ga!#86jv#;`<n20cZ-r{eO6loUc3ZnB~b?8hIK5InWBPlXuJ7DWQ*7DJP+EbGn5u
zCjnV46}ly?ojb})MIpw$T32b(yP8LEX6AtW{q6&(ovM2Cv-nUFJIh65A{CKCy)@X@
zJ5ymaLb#NRT4sC_c8}z|2PLoHnVR=nxYc0r5$0gDtz3vcpCj?bD4u0HzD47Khw*S#
z9uDt@+ya4%)68reZsv>B85{m=>pT=yOBfe#q0G?z9ybL=xg>!o#BN|;-BTakw;+9p
z5yj6E4Q$7ZM~!h@WEz%#euri4n~5W=K8B9o+CbjQtJGF8$pmiTTx0Tu7#f|lrxAgl
zMlv@0LfaGH6+w(}g&QQe)koF7Tm(&B7U9*=ala`gS2f}1NTWI2fg#Rzg+7NDFH;e(
zLO49}>tF74;cS(ejM&Hjh!MbdtP~A=mAvAceliu_sR7}x@w<RR>pELdgIr6rhoVmL
z*a0}3%XNIA0PYXh^MxcfRUIsn5{|^6(YWjU_(#m6A|uR2y;5PQ{%c<X=+JDK6GNt4
z$XvgV!;#Zuq+W>bw)_S+k8X2@!3!)!J#C1O&D(|au%><-BUz&ar|z(~FE(DVFTFcF
zyr+X*qJbGE`L07Zb#}2n5Jq@t^bERCi*N=d0GzIIpAwokeveRY_rRc>7N^Ke=O%O@
z8*m1!VUMG~?&fHu#Pi?=qOVk_PjC%C&VVa+z3sqGpiN>Ie5uOp+YQ+RE`Q_~=`qH8
zKG8m8)TbCHTLh0s+!K<$P(R8<dczBs6Y)FiPSG&=k=`kiYX6<9Po1m*YRRvP(5H!)
z{?(_^6pTrr8}u{~t0FZy%4QeOGk5{P85DViMRa*cTKPx9AzXFik;Qc<e2z&FMLK$_
zi>yk>15>%^^TymlUE%XtgO8nB#O>ndMAzjLBLhb~j|-bsKh)<PThI$Rwiaww(ajXO
zLvPD17=8rE9wuidp97N5olfxT0-?n&uFZ%%f7X}AcUcv6)o8}8g0E?L{shNP_c%qs
zQ`BwSW2?Dr6&5-N*nCDD+w38HQ>Tv4&P3m6m9L9Zw<V0SonvOyu*xx&XgczW8(_;D
zR`bOPFZv{_B1X6IQ3wAl7pDzhC4K}$(YZ5B$%oM*$7SxLZMNXSSM_FEN*tExL<AS7
zwae1GqD|K>#lvS0%r0=M6VZCx>FBkW94)t<oo#t7+B$q&pIxw1D2-c4`h1g~>-dss
z4PQ0WYc)Gxt11Xwixa$E@`r5Ss3B_e>t$7_Wx+|S4eWaiL70-wQgks>k23faQ?qhU
zil2(JH)F-8ac#D{Df~cL{6Sc?31#h)AE|REpc-`>_?n#Da<n&glOuE$eF6;)%B5X<
zu$K$Jqk7GkOOEOWwzr|1Mb&8U6Nb?(;a*?eOOA4x(k^zZ;7eruO00!;R&Q8hguLVB
zEl%{-{$>`qg}r5w7OlHnWWY`QMRYo+b|7=jQIa+G6)i&csC+GVLT2W}$DN;z_0?_6
z8U)x!8#W;yicmX|s>)KUJ4*5&_`p+yJLm$T-ot=ee6^dwe`Y`=WSgQ4p1`F9kwo*+
ztB6jj!2+cyB|384;YjeL2x^b_m7ef$sj(|&MPlmH#0Jys%4hI0gI5@QlEJ4Le38Ld
z82mPa-)Ha}4E}<_w-~(50Q5B>UeMxnDQ<GA$h1KQ+YyB6lVb6T<ZIDTlfxFCKq!qM
zy*`>H_){1V^DtHl(jl#V3IA&d3W*XJe>#nd<ER*NVbTSUVZcZieP6?5s^sANZ^_6r
zh2^7vzQX{F8qt;ODAvv87vs8dc}h8!grE|)$#k?^Gk2L~bHGfQX>-^dHE%IZbKD#<
kZ#IW=W2Rx=h4^-JaQGJUq`pVja_=NQVBVP9THNse09AV$sQ>@~

literal 0
HcmV?d00001

diff --git a/cnn_model.py b/cnn_model.py
new file mode 100644
index 0000000..a648b14
--- /dev/null
+++ b/cnn_model.py
@@ -0,0 +1,274 @@
+"""
+PyTorch 1D CNN Model for Land Use Classification
+Dùng cho 3 channels: NDVI, VH, VV (13 time steps)
+"""
+
+import torch
+import torch.nn as nn
+import torch.optim as optim
+from torch.utils.data import Dataset, DataLoader
+import numpy as np
+from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score, confusion_matrix
+import matplotlib.pyplot as plt
+
+
+class TimeSeriesDataset(Dataset):
+    """Custom Dataset for time series data"""
+    def __init__(self, X, y):
+        self.X = torch.FloatTensor(X)
+        self.y = torch.LongTensor(y)
+    
+    def __len__(self):
+        return len(self.X)
+    
+    def __getitem__(self, idx):
+        return self.X[idx], self.y[idx]
+
+
+class CNN1D(nn.Module):
+    """1D CNN for time series classification
+    
+    Input: (batch_size, 3, 13)
+        - 3 channels: NDVI, VH, VV
+        - 13 time steps
+    Output: (batch_size, num_classes)
+    """
+    def __init__(self, num_classes=8, dropout_rate=0.3):
+        super(CNN1D, self).__init__()
+        
+        # 1D Convolutional layers
+        self.conv1 = nn.Conv1d(in_channels=3, out_channels=32, kernel_size=3, padding=1)
+        self.bn1 = nn.BatchNorm1d(32)
+        self.relu1 = nn.ReLU()
+        self.pool1 = nn.MaxPool1d(kernel_size=2, stride=2)
+        
+        self.conv2 = nn.Conv1d(in_channels=32, out_channels=64, kernel_size=3, padding=1)
+        self.bn2 = nn.BatchNorm1d(64)
+        self.relu2 = nn.ReLU()
+        self.pool2 = nn.MaxPool1d(kernel_size=2, stride=2)
+        
+        self.conv3 = nn.Conv1d(in_channels=64, out_channels=128, kernel_size=3, padding=1)
+        self.bn3 = nn.BatchNorm1d(128)
+        self.relu3 = nn.ReLU()
+        self.pool3 = nn.MaxPool1d(kernel_size=2, stride=2)
+        
+        # Global average pooling
+        self.global_avg_pool = nn.AdaptiveAvgPool1d(1)
+        
+        # Fully connected layers
+        self.fc1 = nn.Linear(128, 64)
+        self.dropout = nn.Dropout(dropout_rate)
+        self.fc2 = nn.Linear(64, num_classes)
+    
+    def forward(self, x):
+        # Conv block 1
+        x = self.conv1(x)
+        x = self.bn1(x)
+        x = self.relu1(x)
+        x = self.pool1(x)
+        
+        # Conv block 2
+        x = self.conv2(x)
+        x = self.bn2(x)
+        x = self.relu2(x)
+        x = self.pool2(x)
+        
+        # Conv block 3
+        x = self.conv3(x)
+        x = self.bn3(x)
+        x = self.relu3(x)
+        x = self.pool3(x)
+        
+        # Global average pooling
+        x = self.global_avg_pool(x)
+        x = x.view(x.size(0), -1)
+        
+        # FC layers
+        x = self.fc1(x)
+        x = self.dropout(x)
+        x = self.fc2(x)
+        
+        return x
+
+
+class CNNTrainer:
+    """Trainer for PyTorch CNN Model"""
+    
+    def __init__(self, num_classes=8, learning_rate=0.001, device=None):
+        self.device = device or torch.device("cuda" if torch.cuda.is_available() else "cpu")
+        self.model = CNN1D(num_classes=num_classes).to(self.device)
+        self.criterion = nn.CrossEntropyLoss()
+        self.optimizer = optim.Adam(self.model.parameters(), lr=learning_rate)
+        self.history = {'train_loss': [], 'val_loss': [], 'train_acc': [], 'val_acc': []}
+        
+        print(f"🚀 Model initialized on device: {self.device}")
+        print(f"   Total parameters: {sum(p.numel() for p in self.model.parameters()):,}")
+    
+    def train_epoch(self, train_loader):
+        """Train one epoch"""
+        self.model.train()
+        total_loss = 0
+        correct = 0
+        total = 0
+        
+        for X_batch, y_batch in train_loader:
+            X_batch, y_batch = X_batch.to(self.device), y_batch.to(self.device)
+            
+            # Forward pass
+            outputs = self.model(X_batch)
+            loss = self.criterion(outputs, y_batch)
+            
+            # Backward pass
+            self.optimizer.zero_grad()
+            loss.backward()
+            self.optimizer.step()
+            
+            # Metrics
+            total_loss += loss.item()
+            _, predicted = torch.max(outputs.data, 1)
+            correct += (predicted == y_batch).sum().item()
+            total += y_batch.size(0)
+        
+        avg_loss = total_loss / len(train_loader)
+        accuracy = correct / total
+        return avg_loss, accuracy
+    
+    def validate(self, val_loader):
+        """Validate model"""
+        self.model.eval()
+        total_loss = 0
+        correct = 0
+        total = 0
+        
+        with torch.no_grad():
+            for X_batch, y_batch in val_loader:
+                X_batch, y_batch = X_batch.to(self.device), y_batch.to(self.device)
+                
+                outputs = self.model(X_batch)
+                loss = self.criterion(outputs, y_batch)
+                
+                total_loss += loss.item()
+                _, predicted = torch.max(outputs.data, 1)
+                correct += (predicted == y_batch).sum().item()
+                total += y_batch.size(0)
+        
+        avg_loss = total_loss / len(val_loader)
+        accuracy = correct / total
+        return avg_loss, accuracy
+    
+    def fit(self, X_train, y_train, X_val, y_val, epochs=50, batch_size=32, verbose=True):
+        """Train model with validation"""
+        train_dataset = TimeSeriesDataset(X_train, y_train)
+        val_dataset = TimeSeriesDataset(X_val, y_val)
+        
+        train_loader = DataLoader(train_dataset, batch_size=batch_size, shuffle=True)
+        val_loader = DataLoader(val_dataset, batch_size=batch_size, shuffle=False)
+        
+        print(f"\n📊 Training start:")
+        print(f"   Train samples: {len(X_train)}")
+        print(f"   Val samples: {len(X_val)}")
+        print(f"   Batch size: {batch_size}")
+        print(f"   Epochs: {epochs}\n")
+        
+        for epoch in range(epochs):
+            train_loss, train_acc = self.train_epoch(train_loader)
+            val_loss, val_acc = self.validate(val_loader)
+            
+            self.history['train_loss'].append(train_loss)
+            self.history['val_loss'].append(val_loss)
+            self.history['train_acc'].append(train_acc)
+            self.history['val_acc'].append(val_acc)
+            
+            if verbose and (epoch + 1) % 10 == 0:
+                print(f"Epoch [{epoch+1}/{epochs}] "
+                      f"Train Loss: {train_loss:.4f}, Acc: {train_acc:.4f} | "
+                      f"Val Loss: {val_loss:.4f}, Acc: {val_acc:.4f}")
+        
+        print(f"\n✅ Training completed!")
+        print(f"   Final Train Acc: {train_acc:.4f}")
+        print(f"   Final Val Acc: {val_acc:.4f}")
+    
+    def predict(self, X_test):
+        """Predict on test data"""
+        self.model.eval()
+        X_test = torch.FloatTensor(X_test).to(self.device)
+        
+        with torch.no_grad():
+            outputs = self.model(X_test)
+            _, predictions = torch.max(outputs, 1)
+        
+        return predictions.cpu().numpy()
+    
+    def evaluate(self, X_test, y_test):
+        """Evaluate on test data"""
+        y_pred = self.predict(X_test)
+        
+        accuracy = accuracy_score(y_test, y_pred)
+        precision = precision_score(y_test, y_pred, average='weighted', zero_division=0)
+        recall = recall_score(y_test, y_pred, average='weighted', zero_division=0)
+        f1 = f1_score(y_test, y_pred, average='weighted', zero_division=0)
+        
+        print(f"\n📈 Test Results:")
+        print(f"   Accuracy:  {accuracy:.4f}")
+        print(f"   Precision: {precision:.4f}")
+        print(f"   Recall:    {recall:.4f}")
+        print(f"   F1-Score:  {f1:.4f}")
+        
+        return {
+            'accuracy': accuracy,
+            'precision': precision,
+            'recall': recall,
+            'f1': f1,
+            'predictions': y_pred,
+            'confusion_matrix': confusion_matrix(y_test, y_pred)
+        }
+    
+    def plot_history(self):
+        """Plot training history"""
+        fig, axes = plt.subplots(1, 2, figsize=(12, 4))
+        
+        # Loss
+        axes[0].plot(self.history['train_loss'], label='Train Loss')
+        axes[0].plot(self.history['val_loss'], label='Val Loss')
+        axes[0].set_xlabel('Epoch')
+        axes[0].set_ylabel('Loss')
+        axes[0].set_title('Training and Validation Loss')
+        axes[0].legend()
+        axes[0].grid(True)
+        
+        # Accuracy
+        axes[1].plot(self.history['train_acc'], label='Train Acc')
+        axes[1].plot(self.history['val_acc'], label='Val Acc')
+        axes[1].set_xlabel('Epoch')
+        axes[1].set_ylabel('Accuracy')
+        axes[1].set_title('Training and Validation Accuracy')
+        axes[1].legend()
+        axes[1].grid(True)
+        
+        plt.tight_layout()
+        plt.show()
+    
+    def save(self, filepath):
+        """Save model"""
+        torch.save(self.model.state_dict(), filepath)
+        print(f"✅ Model saved to {filepath}")
+    
+    def load(self, filepath):
+        """Load model"""
+        self.model.load_state_dict(torch.load(filepath, map_location=self.device))
+        print(f"✅ Model loaded from {filepath}")
+
+
+def reshape_for_cnn(X):
+    """Reshape data for CNN
+    
+    Input: (n_samples, n_features) where n_features = 13*3 = 39 (13 timesteps x 3 channels)
+    Output: (n_samples, 3, 13) - (batch, channels, timesteps)
+    """
+    n_samples = X.shape[0]
+    n_timesteps = 13
+    n_channels = 3
+    
+    # Reshape: (n_samples, 39) -> (n_samples, 3, 13)
+    X_cnn = X.reshape(n_samples, n_channels, n_timesteps)
+    return X_cnn