This is a MIME encoded message
--minaeibi1032462233
Content-Type: text/plain
minaeibi Thu Sep 19 15:03:53 2002 EDT
Added files:
/modules/minaeibi survey4.ppt
Log:
clustering stuff
--minaeibi1032462233
Content-Type: text/plain
Content-Disposition: attachment; filename="minaeibi-20020919150353.txt"
Index: modules/minaeibi/survey4.ppt
+++ modules/minaeibi/survey4.ppt
ࡱ > V U n . yppD!4PNG
IHDR $ PLTE @@@\f bKGD H cmPPJCmp0712 Om IDAThj8 puM/eA2}By]+C_0 M1{?$KKMX,Jc;~YDDM,`~ 'xrLS CqhfX
k%dR2aԔ(m n^`En.hDP' L"D+ ( -0 )MB~)HY
.y8 ^ TG˄E袧 f L_@|SK{ɼ̑y,T ZDZɸeA )S<4HED¨ PK6|5{yG6md-^ jOdc Ů Jys m75 D=R`B&a;1 }~̌-~J%] g;%R>ڌu,].A{6 y$h1zS pq {A Ak 9$k (f~Bn ' aSd6YD@nn ƾPO %q=\qO^ P`~Z` G > ^BCaHP@E(4wRûIy j(lY53(]@/G>/={'3(zf-.5j ڊ#. o-Y
%ȩK rI
E. ]sچ @nvӷE6Z9; wRpz P=+yB~Or
C ʓ{gnc{]_<>p,̘m`u?ԇ8 _# q`BZ]zḀfŘMhՌf1 #{.'f!)\UJ\ag IqRk?2c oɤRjegE}fզD0e߿KL)LGX5 Nn`$"DFvtH?.j{ ɤ.Pv%NR8f.?Aބ,MfRF^P8BRZ >
#ʫ`Mlބ`LLJ11+ Է
VMĴ
ijZM/h
& i<Zi+jG&&DFM.W8ږϣ
7V$WMXP^0`/?mzGx# IENDB` n s 49yPNG
IHDR sBIT|.w PLTE ٟ tRNS 0J bKGD H cmPPJCmp0712 Om |IDAThAkA r {PzQh^Zhz
DO@LA̡BIKBI4[2owYw1];Ǫ|Y_ kŘ˴V@*}~ z}.! [ܽ;ORԻz?Ajt8Uw>?M=p%{R?{C L+ֿL z?s S=,&u?afxw!9 EXUEߊ}3/ڿG>g~f<A<5%'5|+o 7 } #5G7~ 眯Gr+o ~HauָK_οuZ?0o };-C_75Lꚙ?5\)~ ~ph=C[Sg"}ğ_ ψ5?SyX̿EeC#=-\}~7}R=K^OċoG'{?Or'|g<'?\Nii+{Vҫomp? @|wcaB> VY v*H<L3dz_
9a:cu|_뗇xG֊|7#aJ0G;^m7 s &G}9S͛Vu!| 2 ?/b$+ٳI{'yLy ~z??J \YQ#G[/HIy~W <Gh s<t˞d/ys8:=~x@*vNH 8`~ pOѩxnʂ S97 IENDB``!" h쓠'a)|e @ xڥKTQǿ،SD",6mV.^Az|BRlh B6.f)ѶMee̅{z IH0,&rqEֺG7\oM]Vg$93띌;丘 ,/υ%.N|8C ]SJzk?"5#7OW!(}! p1ɮ5/a+Dgt =fAL8_3r ^\/JAaf_rz '|m ZJxVqjWxTfjw?w
aíQ-Xg9hcYw, ުt.˞"뇓V:? Hy RYB/@?<ޙ"Du|[4 IhqXCқ33I@ )b x9tс}yd6\[/Yxe4ۏ֪*)`!Z /Oa߉*sgq ( xڥTMhAvLR"Q, &'
b7=
#B%P<$қ^=y!(zP/^:of}| 3W [(2yQ}لz:qQи'zfcbNW&'Oml?qas rU}_W749tоѹ v Ucm^s pzGC?}±/GGyIy)&>.0~ C?u(҇u4ZbZ/_n`z1z=%[ 4 ~j 5ckPY炻? NoHtC+ߘ9V:b\gy|{P=d|ΙBH2NB 9y{Y
ݒ2/՞Pmvh#dZ\BG[@<A5Ves=Z X ̟ƌv& 2&ATmX3d |p?8{i
/ n oX|9RPNG
IHDR \ ~ Mny bKGD #2 cmPPJCmp0712 ` :IDATx^]-t";N⹑E u}uQJCn]+[ǺֵuQWd-n^ݛd&gO8=%ǐI{s@#;Uނ1̣ r)3 G.̽,
x [/0HG2>=dFpӉm۳[Xl ȝ0XG$>\Kn=;_u q1rtiuY];1:$L1cXYĐ8Ax@>v > \:}Rcưr(5RkM̑uaD&aBqtʣpDn4i2>u+F!)
G&( _@;YTQڕqj
nM /h})ߩeknb3ugX\hݷ;K[/>?{Gd8T`
5V e$oЯV+o٨^owAl*pE>`̵ +=AD<@ ,OaQzto
&>K]:N_kn0E6ʡ H>p9[rX;;<!~
H7&H !FɰO v WHG>LTӽv > Ͱ[-qxq, /52=FC?
|bNhyq&_7;gׂܫ&Htrw I#?*p ,.n R[ɵpDQ|vꏞqkj/ {j7tE;VKH̘o#&g\f9u{fGϸȍTx }m?q܆٤{],[\ ⡶8h# PZGff/4N @n # M'7Ó\&P24 (}*}9~Z2JT(PdYH ʏdp7;oD&W]cS7;ijҐrR do8&I<(*}%S Vڨ))?13*BNhCY. @^hh ~I`r=+} $n @6!C#
|~[~}
%lgjgn @bb *ͳ1goϢH9h d=C YV|+};`t)rJ^{4A^\Ed+m"%KC2(Njf "J_WtIHA2ey # M' ]x`MI9S(n;q^,3F JR6G ]/TA!)%hyoùKRuAI3H+uCSq rw Ub%$`ϴ*_)"Vy x$) 7*!cnh- YB Aܿ $)krrB%nh!J5j {?ܨtCMPz6q5d#bo u9iPrNq=1/կl*.7j? #E8$)ݣhFC@ IF.%m\cDFM~KP #Z )1K(_ O!JL 4$eU 541P74E',Bm
Q}TZ ! Rc> hh\ Se1kϙ@1U Seߑ'wao`G`a+-_w9mL!h,{(T_NhSf+/U*ϸ!郠Z&Un
-524dX}9b
(2|BK
|鏚[" k~顤 ,)3U@ĕ$cM&TZr{+{)Xuت&Iԏ1`V'`BZ;?W`PgD0%0 V E[ +R'dֺA5@$v0%0`ZL4*ޞ#ɉ>}jQ zO3rJgBАZX^D1fD1CXeYt: <{/$#KM6J! (9O1זbsdq70XE|jp^xtPlnrgӤ"&cĢ׃9b
(8rJ!1zmQkP@5?}T}'sja
(xyh=xg\Sf+"[A4@p[4!\S
t6*ɥ 9)1.m\ ,[aYe̢ ׂ,ۮ\[,3I9i4Y66oISBPDĄ*V[z¯B Nuek5TD繞 ɽm4H$BA녖_s}BM.[m$,lUj{+\<li"Nd
n|]2.⡢??NJ{d2@kƶpctJ54]PQ[q92'WjD?plxIڳXks?Yx~K+!.^ DBENǝ4fag20#@_փ2e&5aX} [rl#(jfT
5N;@**s: .
;ݷN&+"!+YR^ҡ\ zq!W1!) KusxB+ 9x+J\ a= SyXeW/ Q= IENDB` (
L
< 6 = E q u a t i o n E q u a t i o n . 3 0 , M i c r o s o f t E q u a t i o n 3 . 0 7 > E q u a t i o n E q u a t i o n . 3 0 , M i c r o s o f t E q u a t i o n 3 . 0 J / 0 | D A r i a l h e t | | w 0 z[ 0 " D G a r a m o n d | | w 0 z[ 0 D T i m e s N e w R o m a n 0 z[ 0 0 D W i n g d i n g s R o m a n 0 z[ 0 @ D T r e b u c h e t M S a n 0 z[ 0 " @ 0.
@ n ? " d d @ @@ `` @ ,
K Q - / 1
5
L $ b $ yppD!4 6 b $ s 49y " 6 2 $ h쓠'a)|e * X 2 $ /Oa߉*sgq b $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ b $ oX|9R 0 A A @ 8 w ʚ; ʚ; g 4 j d j d z[ 0 p p p @ < 4 d d d d @ 0| w < 4 ! d ! d @ < 0| 8 0 _ _ _ P P T 1 0 ? L + 8 S u r v e y , L O N - C A P A D a t a M i n i n g O = #
Clustering $
a !
Partitional Methods k-mean Algorithm Graph Connectivity Nearest Neighbor Method Mixture DensityMaximum Likelihood Estimation Fuzzy k-means clusteringHierarchical MethodsBottom-up approach (agglomerative) Top-down approach (divisive) Z @ Z Z 9 Z Z A Z a a O a & a a a a a a a a a : , 6 Clustering / Partitional Methods / Graph Connectivity 7 7 c & < . 6 Clustering / Partitional Methods / k-mean Algorithm 7 7 c & ? / ; Clustering / Partitional Methods / Nearest Neighbor Method < < # c & c & c & $ 6 T h i s a l g o r i t h m p a r t i t i o n s a s e t o f { x 1 , x 2 , & , x n } p a t t e r n s i n t o a s e t o f k c l u s t e r s { C 1 , C 2 , & , C k } A l g o r i t h m : L e t i = 1 a n d k = 1 . A s s i g n p a t t e r n x 1 t o c l u s t e r C 1 . M a r k p a t t e r n x 1 . i = i + 1 . F i n d t h e n e a r e s t n e i g h b o r o f x i a m o n g t h e p a t t e r n s a l r e a d y a s s i g n e d t o i t s n e a r e s t n e i g h b o r . L e t d j d e n o t e t h e d i s t a n c e b e t w e e n p a t t e r n x i a n d i t s n e a r e s t n e i g h b o r i n c l u s t e r j . I f d j r , t h e n a s s i g n x i t o C j ; o t h e r w i s e k = k + 1 a n d a s s i g n x i t o a n e w c l u s t e r C k . M a r k p a t t e r n x i . I f a l l t h e p a t t e r n s h a v e b e e n m a r k e d , s t o p ; e l s e g o t o s t e p 2 . 6 p $ c c c c c c c c c c c c c c c c c c c c c # c c B c c % c c % c c c c c c c c c c c c c c c c A c c P / R ; - " Clustering / Hierarchical Methods # # c & ` 333 3 ff3 ` 333 3f 33 ff3 ` " 333 3 ̙ ff3 ` K f 3 ̙ ` & e ̙3 g 3f ` f 333 ̙ po7 ` ___ f3 ̙ ;/ f 9 ` ff 3 Lm ` ff 3 LmN Lm > ? " d d @ * ? n A d @ q < n A q F L K #M n ? " d d @ @@ `` P R M ` 2 p > > (
H ? ? " `} X Click to edit Master title style ! ! @
H 0 ? ? " ` R Click to edit Master text stylesSecond levelThird levelFourth levelFifth level ! S
6 # " ` ] `} \ *
6 $ # " ` ` V *
6 # " ` ] `} V *
C @ A B C D E F jJ @ 3 " 0 ` B
s * D jJ " 0 `0 H
0 h ? ___ f3 ̙ ;/ f 9 _ _ _ P P T 1 0 i .
+ D= ' = @B + E d g e 0 (
H c' ? ? " @ ' X Click to edit Master title style ! !
H 06 ? ? " ' [ # Click to edit Master subtitle style $ $
6 L56 # " ` ] `} 6 \ *
6 L:6 # " ` ]} 6 V *
6 >6 # " ` ] `} 6 V *
C @ A B C D E F 8c @ 3 " @ B
s * D jJ " , $
0 H
0 h ? ___ f3 ̙ ;/ f 9 _ _ _ P P T 1 0 i .
+ ity D= ' = @B +
0 z r P
(
0 6 P
6 P *
0 h6 6 R *
d
c $ ?
6
0 l6
0 6 R Click to edit Master text stylesSecond levelThird levelFourth levelFifth level !
S
6 ' _ P 6 P *
6 X' _ 6 R *
H
0 h ? 33 8 0 _ _ _ P P T 1 0 . 9\ K0 , < (
, ~
, s * y' `} ' ~
, s * z' ` ' H
, 0 h ? f 3 | y _ _ _ P P T 1 0 Y + D= ' = @B + } u (
r
S ' `} '
P A M S T @ 0H '
< ' P
W An algorithm used by Zahn uses the MST to determine structures in the graph of the data set. The idea is that the data points within the same cluster are closer to each other than to data points in different clusters. 1. Construction of the minimum spanning tree with n data points. 2. Identify inconsistent edges in the minimum spanning tree. 3. Remove c-1 inconsistent edges. The remaining connected components are the clusters f 0
a 2 W L H
0 h ? ___ f3 ̙ ;/ f 9 8 0 _ _ _ P P T 1 0 . \pyI d
d (
x
c $ ' `} ' P
c $ ' 0v ' T h e m e a n v a l u e o f a c l u s t e r i s c o m p u t e d a c c o r d i n g t o t h i s f u n c t i o n : T h e t o t a l s q u a r e d - e r r o r i s c o m p u t e d i n t h i s w a y : 1 . C h o o s e a n i n i t i a l p a r t i t i o n w i t h k < n c l u s t e r s 2 . G e n e r a t e a n e w p a r t i t i o n b y a s s i g n i n g a p a t t e r n t o i t s n e a r e s t c l u s t e r c e n t e r . 3 . R e - c o m p u t e n e w c l u s t e r c e n t e r s . 4 . G o t o s t e p 2 u n l e s s n o c h a n g e i n c l u s t e r c e n t e r s . 5 . R e t u r n , , & , a s t h e m e a n v a l u e s o f C 1 , C 2 , & , C k . y w c $ c c c c c c c
0
Z A 6 ? A A ? @ ` 6 '
0
Z A 7 ? A A ? 0 7 ' H
0 h ? ___ f3 ̙ ;/ f 9 8 0 _ _ _ P P T 1 0 . \pyI @ P @ (
x
c $ 6 `} 6
S ' ` 6 " hu= H
0 h ? ___ f3 ̙ ;/ f 9 8 0 _ _ _ P P T 1 0 . \pyI ?
? (
x
c $ d]' `} '
` A h i r a r c h i c a l 0 \ '
r A * t a b e l l e _ s i m i l a r i t i e s '
< _' p 0t : Matrix of an data set showing the distances or similarities, respectively, of the clusters. The marked element in the left matrix shows the minimal distance between two clusters that are joined to one cluster in the right matrix. 6 0 c w
< j' Dentograms for the single linkage, left, and the complete linkage, right, are displayed. It shows the distance level where the grouping took place and between which clusters. & c
H
0 h ? ___ f3 ̙ ;/ f 9 8 0 _ _ _ P P T 1 0 . \pyI o xXMLSA}< ŖшL`s4&4-M84B =p@O&&/ ^1Ez0 Ř@uZJٝfަ/^ߺ_0Hgak#n M3eϔ询.g1|_D>%/3gToP8('"\ 8bGPiW \~9+& 7L, 5Sa2c~/rw> $ rqu湻jhʐe/ѿKw~[*:P ]~㰰9pT ;+!@\
ك\0]a% myު!Ԫ-xh0tnCMPaH)D QBHOUd9iR 1 .U5#)]!jq΄F{CX&?S9T܇]C4ȸTC/4*{P(`õJB|7?4T哈ZOtYC&?f=fgaT~]UhS~ĹV/U~geoO$ q`4m%/E KXҢQv6UfF-~[,} F<X?q>6Gn~Ia|},_Q)S F3MB's Nt6^ 3 a7tf;Tɪz_* )B-h J#(Q{̍3_=pB<W,^T{MW2zmoV3D} s"q5Ÿru"'=셣^x?yYyIZ\t܌Xc3Kd k xXOHQ7iM`tdm-]Z`xa:Y^ȼDv&إKmjBX,73o73^z|,:l\ȰL2 mR)k<U!FaC 9l(lKU`OA:.ϩLƵSa'9kp <Z<\>9'~e/HA*
-V>C(P[{ɽ?5EN=^dO :
9E1&.&.!G\
<^rJf y:CPK_hc62RBrBKB!$16wگi7#氣jH#idScb8 %uToλA+` 6fZ1 I_qnC=!oC(HD ~ /JL_UB Y\u !{|_6 4WGjt<EK=vuoުu$/)@z!0Ǣ~ȯuzzޓ֧e}i\/gWyu0N827 'MvY NvZc3!M֛sՓ oG!ݝw9 i
h pZ¨5h 2[7gT,=
1qh+ ܞ ԁi"p ǛKtTwiʓFCI f$:CÑ(ȷ-|ѳ 7 7m1 u'\y\ znhw1 r< - % 3 : ` N6 E c; ^K N B - ?
HR ? Oh +'0 ` h p
$ , 1 Data Mining for an Educational Web-based system Poi rje Edge Min rje 10 Microsoft PowerPoint duc@ Y, @ D{8\@ %$ ] d G , g 4 - - @ ! - - ' @ Garamond - . 2
< 9/15/2002 . "System - @ Garamond - . 2
Survey, LON
. - @ Garamond - . 2
- . - @ Garamond - . 2
CAPA Data Mining . - @ Garamond - . 2
1 . - ̙ - -
% ) [ ) - - ' ̙ - - % 2 - - ' @ Times New Roman - . f3 2
[ <
Clustering' . - @ Wingdings - . ̙ 2
< n . - @ Times New Roman - . 2
b Partitional . - @ Times New Roman - . 2
Methodse% . - @ Wingdings - . ;/ 2
b q . - @ Times New Roman - . 2
k . - @ Times New Roman - . 2
- . - @ Times New Roman - . 2
mean Algorithm
. - @ Wingdings - . ;/ 2
b q . - @ Times New Roman - . $ 2
Graph Connectivity
. - @ Wingdings - . ;/ 2
?b q . - @ Times New Roman - . + 2
? Nearest Neighbor Method
. - @ Wingdings - . ;/ 2
nb q . - @ Times New Roman - . 2
n Mixture Density
. - @ Wingdings - . ;/ 2
b q . - @ Times New Roman - . 4 2
Maximum Likelihood Estimation
. - @ Wingdings - . ;/ 2
b q . - @ Times New Roman - . 2
Fuzzy . - @ Times New Roman - . 2
k . - @ Times New Roman - . 2
- . - @ Times New Roman - . 2
means clustering
. - @ Wingdings - . ̙ 2
< n . - @ Times New Roman - . % 2
b Hierarchical Methods % . - @ Wingdings - . ;/ 2
5b q . - @ Times New Roman - . 2
5 Bottom
. - @ Times New Roman - . 2
5 - . - @ Times New Roman - . 2
5 up approach ( . - @ Times New Roman - . 2
5 agglomerative
. - @ Times New Roman - .
2
5 ) . - @ Wingdings - . ;/ 2
db q . - @ Times New Roman - . 2
d Top . - @ Times New Roman - . 2
d - . - @ Times New Roman - . 2
d down approach ( . - @ Times New Roman - . 2
d divisive
. - @ Times New Roman - . 2
d, ) . - ՜. +,0
n On-screen Show msu Fl ( A
Arial Garamond Times New Roman
Wingdings Trebuchet MS Edge Microsoft Equation 3.0 Clustering 7 Clustering / Partitional Methods / Graph Connectivity 7 Clustering / Partitional Methods / k-mean Algorithm < Clustering / Partitional Methods / Nearest Neighbor Method # Clustering / Hierarchical Methods Fonts Used Design Template Embedded OLE Servers Slide Titles _R rje r j e
! " # $ % & ' ( ) * + , - . / 0 1 2 3 4 5 6 8 9 : ; < = > ? @ A B C D F G H I J K L N O P Q R S T W R o o t E n t r y dO ) P i c t u r e s C u r r e n t U s e r M S u m m a r y I n f o r m a t i o n ( 7 P o w e r P o i n t D o c u m e n t ( R D o c u m e n t S u m m a r y I n f o r m a t i o n 8 E
--minaeibi1032462233--