240 lines (236 with data), 22.7 kB

PK
Îl‹K	META-INF/þÊPK
Íl‹KøÈ÷wüüMETA-INF/MANIFEST.MFManifest-Version: 1.0
Ant-Version: Apache Ant 1.9.4
Created-By: 1.8.0_91-b15 (Oracle Corporation)
Class-Path: lib/weka.jar lib/liblinear-1.92.jar lib/liblinear-java-1.9
 6-SNAPSHOT.jar
X-COMMENT: Main-Class will be added automatically by build

PK
Îl‹Kmain/PK
Îl‹K
main/java/PK
Îl‹Kweka/PK
Îl‹Kweka/classifiers/PK
Îl‹Kweka/classifiers/functions/PK
Îl‹K åþ5T5T*weka/classifiers/functions/LibLINEAR.classÊþº¾4•

#
å	

æ	

ç	
è
é	

ê	

ë?PbMÒñ©ü	

ì?¹™™™™™š	

í	

î	

ï	

ð	

ñ	

ò	

ó	

ô	

õ	

ö
÷

å
ø

ù


ú

û

û
ü	
ý
þ

ÿ	

					

+
å

-
+
1 !"#$%&'
<(

#)*+*,
+-
./
x01
23	
4
Q5

6

è7
89

:

;
.<

=

>

?

@

A

B

#CD
b
åEFGH
IJ

K
LM

NO

P

Q

RS

TU

VW

XFYZF[
h
Q\
Q]

^

è_

`abcdefghi
‡j
‡k
‡lmno
Žp
Žqrs
‘tu
”
å	”v	”w	”x	”y	”zå{å|å}å~
Ÿ€å	
‚
΃
΄
Î…	
†
Òƒ
Ò„
Ò…
‡ƒ
‡„
‡…

ˆå‰åŠ
‹Œ	
è	
èŽ	
菐‘
´t
’“
”•
”–

#—
˜™	š›
˜œ	š	šž	šŸ	š 
˜¡	š¢

裤
¥¦	š§

—
˜¨©
˪
Ë«¬
Î
å
έ
‡®¯
Ò
å
Ò­
Ë°å±
˲³´
Øtµ
Û
å
Û¶
‡­
ËŠ
ˉ	
·
˸¹º¹»¼
º	Ÿ½
¾¿	
À
”Á
”Â
”Ã

Ä

Å
Ӯ
ËÇ	
È

ÉÊË
ô
å
’ÌÍÎ
ôÏÐ
‹Ñ
ˁ
Ë}ÒÓ
¾Ô

Õ
.Ö×Ø
ËÙ
‹ÚÛ
’PÜ
ô
ûÝÞ


å

ßàá

âã	
èäå	
èæç	
èèé	
èêëìí	
èîï	
èðñ	
èòóôõ
REVISION
Ljava/lang/String;

ConstantValue
serialVersionUID

J
½9
m_Model
Lde/bwaldvogel/liblinear/Model;
m_Filter
Lweka/filters/Filter;
m_Normalize

Z
TAGS_SVMTYPE
[Lweka/core/Tag;
DEFAULT_SOLVER
$Lde/bwaldvogel/liblinear/SolverType;
m_SolverType
m_eps

D
	m_epsilon
m_Cost
m_MaxIts

I
m_Bias

m_WeightLabel
[I
m_Weight
[D
m_ProbabilityEstimates
m_NominalToBinary
5Lweka/filters/unsupervised/attribute/NominalToBinary;
m_ReplaceMissingValues
:Lweka/filters/unsupervised/attribute/ReplaceMissingValues;
m_Header
Lweka/core/Instances;
m_Counts
m_x1
m_x0
<init>
()V
Code
LineNumberTable
LocalVariableTable
this
&Lweka/classifiers/functions/LibLINEAR;
getModel
!()Lde/bwaldvogel/liblinear/Model;

globalInfo
()Ljava/lang/String;
getTechnicalInformation
"()Lweka/core/TechnicalInformation;
result
 Lweka/core/TechnicalInformation;
listOptions
()Ljava/util/Enumeration;
Ljava/util/Vector;
en
Ljava/util/Enumeration;
LocalVariableTypeTable
&Ljava/util/Vector<Ljava/lang/Object;>;

StackMapTableö

setOptions
([Ljava/lang/String;)V
options
[Ljava/lang/String;
tmpStrZ

Exceptions

getOptions
()[Ljava/lang/String;
Ljava/util/List;
$Ljava/util/List<Ljava/lang/String;>;÷

setSVMType
(Lweka/core/SelectedTag;)V
value
Lweka/core/SelectedTag;

setSolverType
'(Lde/bwaldvogel/liblinear/SolverType;)V

solverType

getSolverType
&()Lde/bwaldvogel/liblinear/SolverType;

getSVMType
()Lweka/core/SelectedTag;
SVMTypeTipText
getEpsilonParameter
()D
setEpsilonParameter
(D)V

v
epsilonParameterTipText
getMaximumNumberOfIterations
()I
setMaximumNumberOfIterations
(I)V
 maximumNumberOfIterationsTipText
setCost
getCost
costTipText
setEps
getEps

epsTipText
setBias
getBias
biasTipText
normalizeTipText
setNormalize
(Z)V
getNormalize
()Z

setWeights
(Ljava/lang/String;)V

i

weightsStr
tok
Ljava/util/StringTokenizer;h

getWeights
sb
Ljava/lang/StringBuilder;
÷
weightsTipText
setProbabilityEstimates
getProbabilityEstimates
probabilityEstimatesTipText

getParameters
%()Lde/bwaldvogel/liblinear/Parameter;
	parameter
#Lde/bwaldvogel/liblinear/Parameter;o

getProblem
M([[Lde/bwaldvogel/liblinear/FeatureNode;[DI)Lde/bwaldvogel/liblinear/Problem;
vx
'[[Lde/bwaldvogel/liblinear/FeatureNode;
vy
	max_index
problem
!Lde/bwaldvogel/liblinear/Problem;
instanceToArray
<(Lweka/core/Instance;)[Lde/bwaldvogel/liblinear/FeatureNode;
idx
val
instance
Lweka/core/Instance;
count
nodes
&[Lde/bwaldvogel/liblinear/FeatureNode;
index
distributionForInstance
(Lweka/core/Instance;)[D

k
labels
prob_estimates

prediction

x
A
?
getCapabilities
()Lweka/core/Capabilities;
Lweka/core/Capabilities;ø
buildClassifier
(Lweka/core/Instances;)V
z0
z1
inst

classValue

d
insts
y0
y1ùÝ©
­¼
toString

j
numNonEmptyClasses
start
Ljava/lang/StringBuffer;

wË
getRevision
main
args
<clinit>

SourceFile
LibLINEAR.java
L
M
.
/
0
1úû
5
4
5
6
5
7
8
9
8
:
8
;
<
=
8
>
?
@
A
B
1
J
8
K
8
,
-
java/lang/StringBuilder
.A wrapper class for the liblinear classifier.
üý
W
X
Ø
V
weka/core/TechnicalInformationþ

L
SRong-En Fan and Kai-Wei Chang and Cho-Jui Hsieh and Xiang-Rui Wang and Chih-Jen Lin	

5LIBLINEAR - A Library for Large Linear Classification
2008
,http://www.csie.ntu.edu.tw/~cjlin/liblinear/

IThe Weka classifier works with version 1.95 of the Java port of LIBLINEAR
java/util/Vector
weka/core/Option
ê	Set type of solver (default: 1)
	for multi-class classification
		 0 -- L2-regularized logistic regression (primal)
		 1 -- L2-regularized L2-loss support vector classification (dual)
		 2 -- L2-regularized L2-loss support vector classification (primal)
		 3 -- L2-regularized L1-loss support vector classification (dual)
		 4 -- support vector classification by Crammer and Singer
		 5 -- L1-regularized L2-loss support vector classification
		 6 -- L1-regularized logistic regression
		 7 -- L2-regularized logistic regression (dual)
	for regression
		11 -- L2-regularized L2-loss support vector regression (primal)
		12 -- L2-regularized L2-loss support vector regression (dual)
		13 -- L2-regularized L1-loss support vector regression (dual)

S
-S <int>
L
(	Set the cost parameter C
	 (default: 1)

C
-C <double>
3	Turn on normalization of input data (default: off)
-Z
®	Use probability estimation (default: off)
currently for L2-regularized logistic regression, L1-regularized logistic regression or L2-regularized logistic regression (dual)! 

P
-P
8	Set tolerance of termination criterion (default: 0.001)

E
-E <double>
>	Set the parameters C of class i to weight[i]*C
	 (default: 1)

W
-W <double>
T	Add Bias term with the given value if >= 0; if < 0, no bias term added (default: 1)

B
-B <double>
K	The epsilon parameter in epsilon-insensitive loss function.
	(default 0.1)

L
-L <double>
=	The maximum number of iterations to perform.
	(default 1000)
-I <int>
[
\ö
•
\
„
weka/core/SelectedTag
2
3
L
q
r
„ 
ˆ
€
‹
€!"
’
“
Ž
€
–
—
¢
“

€
…
†
e
f
java/util/ArrayList
-S÷#$
ü%
-C
‰
~ü&
-E
Œ
~
-B

~
”
•

V
-W
£
•
-L
}
~
-I
ƒ
„'
„
java/lang/String()*+,-.
„/0
u
v
The type of SVM to use.
?The epsilon parameter of the epsilon insensitive loss function.
,The maximum number of iterations to perform.
The cost parameter C.
+The tolerance of the termination criterion.
aIf >= 0, a bias term with that value is added; otherwise (<0) no bias term is added (default: 1).
Whether to normalize the data.
java/util/StringTokenizer

 
L12
„3
V
BThe weights to use for the classes, if empty 1 is used by default.
Whether to generate probability estimates instead of -1/+1 for classification problems (currently for L2-regularized logistic regression only!)
!de/bwaldvogel/liblinear/Parameter
L4
–5
"java/lang/IllegalArgumentException
vx and vy must have same size
L
—
de/bwaldvogel/liblinear/Problem6
<7
<8
8
Â9:
A;
„
»<=
„>?
#de/bwaldvogel/liblinear/FeatureNode
L@A
„
E
FBCD
•EF
C
DG
²
³H
„IJKL
•M
5N
5O
5
weka/core/WekaException
Yprobability estimation is currently only supported for L2-regularized logistic regressionPQRSTUVW
Å
ÆøX
MY[\]^_\`\a\b\c^d\e
„
Solver üf
 is not supported!g\h
Ê
weka/core/Instances
L
Êi
M
8weka/filters/unsupervised/attribute/ReplaceMissingValuesjklm
3weka/filters/unsupervised/attribute/NominalToBinary
¶n
Î
~o
„
java/lang/Exception
LAll class values are the same. At least two class values should be different
-weka/filters/unsupervised/attribute/Normalizep
“
I
Aqrùs
•t
weka/core/Instance
»
<uvwx
1y
Mz
M{
M
ª
«
¥
¦|}
L~
G
H
S
T
LibLINEAR: No model built yet.
java/lang/StringBuffer€
LibLINEAR wrapper




ü
Model for class 
s‚

+

-ƒ„…
„†‡
 * 

(normalized) ˆ‰Š
V



'NOTE: CLASS HAS ALSO BEEN NORMALIZED.


$weka/classifiers/functions/LibLINEAR
1.9.0‹Œ

weka/core/Tag
+L2-regularized logistic regression (primal)
L
;L2-regularized L2-loss support vector classification (dual)Ž
5
=L2-regularized L2-loss support vector classification (primal)
5
;L2-regularized L1-loss support vector classification (dual)
5
3support vector classification by Crammer and Singer‘
5
4L1-regularized L2-loss support vector classification
"L1-regularized logistic regression
)L2-regularized logistic regression (dual)’
5
9L2-regularized L2-loss support vector regression (primal)“
5
7L2-regularized L2-loss support vector regression (dual)”
5
7L2-regularized L1-loss support vector regression (dual)
#weka/classifiers/AbstractClassifier
%weka/core/TechnicalInformationHandler
java/util/Enumeration
java/util/List
weka/core/Capabilities
java/util/Iterator
"de/bwaldvogel/liblinear/SolverType
L2R_L2LOSS_SVC_DUAL
append
-(Ljava/lang/String;)Ljava/lang/StringBuilder;
#weka/core/TechnicalInformation$Type
Type
InnerClasses
MISC
%Lweka/core/TechnicalInformation$Type;
((Lweka/core/TechnicalInformation$Type;)V
$weka/core/TechnicalInformation$Field
Field
AUTHOR
&Lweka/core/TechnicalInformation$Field;
setValue
;(Lweka/core/TechnicalInformation$Field;Ljava/lang/String;)V
TITLE
YEAR
URL
NOTE
:(Ljava/lang/String;Ljava/lang/String;ILjava/lang/String;)V

addElement
(Ljava/lang/Object;)V
hasMoreElements
nextElement
()Ljava/lang/Object;
elements
weka/core/Utils
	getOption
((C[Ljava/lang/String;)Ljava/lang/String;
length
java/lang/Integer
parseInt
(Ljava/lang/String;)I
(I[Lweka/core/Tag;)V
getId
java/lang/Double
parseDouble
(Ljava/lang/String;)D
getFlag
(C[Ljava/lang/String;)Z
add
(Ljava/lang/Object;)Z
(I)Ljava/lang/StringBuilder;
(D)Ljava/lang/StringBuilder;
size
toArray
(([Ljava/lang/Object;)[Ljava/lang/Object;
getTags
()[Lweka/core/Tag;
getSelectedTag
()Lweka/core/Tag;
getID
getById
'(I)Lde/bwaldvogel/liblinear/SolverType;
'(Ljava/lang/String;Ljava/lang/String;)V
countTokens
	nextToken
+(Lde/bwaldvogel/liblinear/SolverType;DDID)V
([D[I)V

l

n
bias
#[[Lde/bwaldvogel/liblinear/Feature;

y
	numValues
(I)I

classIndex
valueSparse
(I)D
(ID)V

numAttributes
input
(Lweka/core/Instance;)Z

batchFinished
output
()Lweka/core/Instance;
weka/filters/Filter

numClasses
classAttribute
()Lweka/core/Attribute;
weka/core/Attribute
	isNominal
L2R_LR
L2R_LR_DUAL
L1R_LR
de/bwaldvogel/liblinear/Model
	getLabels
()[I
de/bwaldvogel/liblinear/Linear
predictProbability
F(Lde/bwaldvogel/liblinear/Model;[Lde/bwaldvogel/liblinear/Feature;[D)D
predict
D(Lde/bwaldvogel/liblinear/Model;[Lde/bwaldvogel/liblinear/Feature;)D

disableAll
!weka/core/Capabilities$Capability

Capability
NOMINAL_ATTRIBUTES
#Lweka/core/Capabilities$Capability;
enable
&(Lweka/core/Capabilities$Capability;)V
NUMERIC_ATTRIBUTES
DATE_ATTRIBUTES
MISSING_VALUES

NOMINAL_CLASS
enableDependency

NUMERIC_CLASS
ordinal
-(Ljava/lang/Object;)Ljava/lang/StringBuilder;
MISSING_CLASS_VALUES
testWithFail
deleteWithMissingClass
setInputFormat
(Lweka/core/Instances;)Z
	useFilter
A(Lweka/core/Instances;Lweka/filters/Filter;)Lweka/core/Instances;
(I)Lweka/core/Instance;
numInstances
setIgnoreClass
iterator
()Ljava/util/Iterator;
hasNext
next
java/lang/Math
max
(II)I
m_Debug
disableDebugOutput
enableDebugOutput
resetRandom
train
e(Lde/bwaldvogel/liblinear/Problem;Lde/bwaldvogel/liblinear/Parameter;)Lde/bwaldvogel/liblinear/Model;
(Lweka/core/Instances;I)V
getFeatureWeights
()[D
,(Ljava/lang/String;)Ljava/lang/StringBuffer;
(I)Ljava/lang/String;
abs
(D)D
getNumDecimalPlaces
doubleToString
(DII)Ljava/lang/String;
	attribute
(I)Lweka/core/Attribute;
name

runClassifier
3(Lweka/classifiers/Classifier;[Ljava/lang/String;)V
(ILjava/lang/String;)V
L2R_L2LOSS_SVC
L2R_L1LOSS_SVC_DUAL
MCSVM_CS
L1R_L2LOSS_SVC
L2R_L2LOSS_SVR
L2R_L2LOSS_SVR_DUAL
L2R_L1LOSS_SVR_DUAL!

#

$
%
&

'

(
)

'
*
,
-
.
/
0
1
2
3
4
5
6
5
7
8
9
8
:
8
;
<
=
8
>
?
@
A
B
1
C
D
E
F
G
H
I
A
J
8
K
8.

L
M

N¼
Z*·
*
µ*µ*²µ**´µ*µ	*
µ*µ
*èµ*µ*¼
µ*¼µ*µ*µ*µ±
O>–«	°ÁÆË$Ð+Õ0Ú7ß<áCãJéO
T

P
Z
Q
R

S
T

N/

*´°
O
¥
P

Q
R

U
V

ND
»Y·¶*¶¶¶¶°
O



P

Q
R

W
X

N†:»Y²·L+² !¶"+²#$¶"+²%&¶"+²'(¶"+²)*¶"+°
O



&
/
8
!
P:
Q
R/
Y
Z


[
\

N
`Ê»+Y·,L+»-Y./0·1¶2+»-Y345·1¶2+»-Y678·1¶2+»-Y9:;·1¶2+»-Y<=>·1¶2+»-Y?@A·1¶2+»-YBCD·1¶2+»-YEFG·1¶2+»-YHIJ·1¶2*·KM,¹L
™+,¹M
¶2§ÿí+¶N°
O:
,
.
>,
@>
BP
Fb
Ht
J†
L˜
Pª
T¯
U¸
VÅ
Y
P Ê
Q
RÂ
Y
]
¯
^
_
`
Â
Y
a

bý¯
c
d

e
f
N
¿øS+¸OM,¶P™*»QY,¸R²S·T¶U§*»QY*´¶V²S·T¶UC+¸OM,¶P™*,¸W¶X§*¶XE+¸OM,¶P™*,¸W¶Y§
*¶Y*Z+¸Z¶[B+¸OM,¶P™*,¸W¶\§*¶\*W+¸O¶]*P+¸Z¶^L+¸OM,¶P™*,¸W¶_§
*
¶_I+¸OM,¶P™*,¸R¶`§
*è¶`*+·a±
Ov
¡
¢
£#
¥8
¨?
©F
ªQ
¬V
¯]
°d
±o
³v
¶€
¸‡
¹Ž
º™
¼ž
¿¨
Á²
ù
ÄÀ
ÅË
ÇÒ
ÊÙ
Ëà
Ìë
Îò
Ñ÷
Ò
P ø
Q
Rø
g
h
ñ
i
&
bü#
j",
k
Ø

l
m

N
ø
R»bY·cL+d¹eW+»Y·f¶*´¶V¶g¶¹eW+h¹eW+»Y·f¶*¶i¶j¶¹eW+k¹eW+»Y·f¶*¶l¶j¶¹eW+m¹eW+»Y·f¶*¶n¶j¶¹eW*¶o™+8¹eW*¶p¶P™)+q¹eW+»Y·f¶*¶p¶¶¹eW*¶r™+;¹eW+s¹eW+»Y·f¶*¶t¶j¶¹eW+u¹eW+»Y·f¶*¶v¶g¶¹eW++¹w
½x¹yÀz°
OV
Û
Ý
Þ1
à:
áW
ã`
ä}
æ†
ç£
éª
ê³
í½
îÆ
ïã
òê
óó
öü
÷

ù
"
ú
?
ü
P
R
Q
R
J
g
n

`

J
g
o

b
ü³
p/

q
r

N^+¶{²S¦*+¶|¶}¸~¶±
O

P
Q
R
s
t

b

u
v

N>*+µ±
O

P
Q
R
w
5

x
y

N/

*´°
O

P

Q
R

z
{

N<
»QY*´¶V²S·T°
O

P

Q
R

|
V

N-

€°
O
"
P

Q
R

}
~

N/
*´¯
O
,
P

Q
R


€

N>*'µ±
O
67
P
Q
R

8


‚
V

N-

°
O
@
P

Q
R

ƒ
„

N/

*´¬
O
I
P

Q
R

…
†

N>*µ±
O
RS
P
Q
R

<


‡
V

N-

‚°
O
\
P

Q
R

ˆ
€

N>*'µ
±
O
ef
P
Q
R
s
8


‰
~

N/
*´
¯
O
n
P

Q
R

Š
V

N-

ƒ°
O
x
P

Q
R

‹
€

N>*'µ	±
O
‚
P
Q
R
s
8


Œ
~

N/
*´	¯
O
Š
P

Q
R


V

N-

„°
O
”
P

Q
R

Ž
€

N>*'µ±
O
ž
P
Q
R
s
8



~

N/
*´¯
O
¦
P

Q
R


V

N-

…°
O
°
P

Q
R

‘
V

N-

†°
O
º
P

Q
R

’
“

N>*µ±
O
ÃÄ
P
Q
R
s
1


”
•

N/

*´¬
O
Ì
P

Q
R

–
—

N»E»‡Y+ˆ·‰M*,¶Š¼µ*,¶Š¼
µ>*´¾¢*´,¶‹¸WR*´O„
§ÿà±
O"Ö×ØÛ*Ü7Ý>ÛDß
P*!#
˜
<E
Q
RE
™
&
:
š
›
bý!
œ
ú"


V

N4»Y·L=*´¾¢ž
+ˆ¶W+*´1¶jW„
§ÿÞ+¶°
Oéêëìî)ê/ñ
P 
%
˜
<4
Q
R,
ž
Ÿ

b
ý

 
ú

¡
V

N-

Ρ
O
û
P

Q
R

¢
“

N>*µ±
O

P
Q
R
s
1


£
•

N/

*´¬
O

P

Q
R

¤
V

N-

°
O

P

Q
R
¥
¦

N€
2»ŽY*´*´
*´	*´*´·L*´¾ž+*´*´¶+°
O$&$'0*
P2
Q
R
§
¨

b
ü0
©
ª
«

Nº?+¾,¾Ÿ
»‘Y’·“¿»”Y·•:,¾µ–µ—*¶nµ˜+µ™,µš°
O&	78;=!>'?0@6A<C
P4?
Q
R?
¬
­
?
®
A?
¯
<%
°
±
b

²
³
N
¾	À=>+¹›
¢++¹œ+¹
 §+¹ž—™„
„
§ÿÑ*´—›„
½ŸN66+¹›
¢I+¹œ6+¹ž9+¹
 §!—š§-»ŸY`· S„
„
§ÿ²*´—›-»ŸY+¹¡
`*´· S-°
ObOQRS!U-V0Q6Z?[B_G`JaXcbdlfwgzij„m•n˜ažr§s¾v
P\	2
˜
<b6
´
<l,
µ
8MQ
˜
<À
Q
RÀ
¶
·
¾
¸
<Gy
¹
ºJv
»
<
b"ý

úþ
æ

ý,
	ùú
k
Ø

¼
½
N

*´¢+¶£W*´¢¶¤W*´¢¶¥L*´¦+¶§W*´¦¶¨W*´¦¶©L*´Æ*´+¶ªW*´¶«W*´¶¬L*+¶­M+¹®
¼N+¹¯
¶°™p*´™i*´²±¥!*´²²¥*´²³¥
»´Yµ·¶¿*´¶·:+¹®
¼:*´,¸¸X6¾¢-.1R„
§ÿé§1*´,¸¹9+¹¯
¶°™-ŽR§-*´k*´cR-°
On‚	ƒ„†"‡*ˆ2Š9‹BŒJRX‘a’t“’”œ—¥˜¯šºÅžÑ× Ú¡ä¢ð£ù¥
ª
PR½
¾
<¥2
¿
?¯(
À
Aä$
Á
8


Q
R


¶
·
X²
Â
ºa©
Y
A
b#ûRýIæ
Ãþ 
Ä
Ã
øüú
k
Ø

Å
Æ

N
¡*·ºL+¶»+²¼¶½+²¾¶½+²¿¶½+²À¶½+²Á¶Â+²Ã¶Â*´¶Ä›*´¶Ä£
+²Á¶½§H*´¶Ä¡*´¶Ä
£
+²Ã¶½§&»‘Y»Y·Ŷ*´¶ÆǶ¶·“¿+²È¶½+°
O>³´	·¸¹º%½,¾3ÁIÂSÃkÄuƘȟÉ
P¡
Q
Rœ
Y
Ç

b
üS
È!"

É
Ê
NÎ
ü*
µ¦*
µ*¶É+¶Ê»ËY+·ÌL+¶Í*»ÎY·Ïµ¢*´¢+¶ÐW+*´¢¸ÑL*»ÒY·Óµ¦*´¦+¶ÔW+*´¦¸ÑL+¶Õ¹Ö
I6+¶×¢+¶Õ¹Ö
(—š	„
§ÿä+¶× 
»ØYÙ·Ú¿+¶Õ¹Ö
9*¶o™+*»ÛY·Üµ*´ÀÛ¶Ý*´+¶ÞW+*´¸ÑL*´Æ:+¶Õ¹Ö
9+¶Õ¹Ö
9	*(g	goµ*(*´kgµ§
*µ*µ+¶ß¶°™=*+¶à¼µá+¶â:¹ã
™#¹ä
Àå:*´á¹Ö
Ž\1cR§ÿÙ+¶×¼:+¶×½æ:6	6

+¶×¢H+
¶Õ:*¶­:¾ž	¾d2´ç¸è6	
S¹Ö
9


R„

§ÿµ*´éš	¸ê§¸ë¸ì**	¶í*¶î¸ïµ*»ËY+·ðµñ±
OÞ7ÔÕ
×ÚÛÝ*Þ3ß<áGâPãYædçgè€é†ëî™ð¦ò­ó¸ôÃõÌöÕùÜúèûõü
ý
þ
ÿ


'
1
M
^
a

i
r
u

‰
‘
—
©
°
¹
À
Æ
Í
Ó
Ö"
Ù%
î(
û)
P˜è(
Ë
8õ
Ì
8	
M
Í
·
‰7
Í
·
‘/
Â
º
¹
Î
8

xN
Ï
<

ü
Q
R
ü
Ð
H
d
˜
Ñ
8g
•
»
<¦
V
Ò
8
i“
®
A
rŠ
¬
­
u‡
¯
<	
b>
ýg
ü;=	ü
Óú)ÿ	
Ô
Õ

Ã
Ö

ý0
×æø
k
Ø

Ø
V

NÔþ*¶òÇó°»ôY·õL*¶ò¶öM+»Y·÷¶*¶ò¶Æø¶¶¶ùW*´ñ¶ß¶°™
·>6*´á¾¢*´á1—ž„
„
§ÿä66*´ñ¶à §
*´ñ¶à¢
o*´á*´ñ¶à §1—ž
N+»Y·ú¶*´ñ¶à *´ñ¶ß¶û§*´ñ¶ß¶û¶ø¶¶¶ùW„
66*´ñ¶ü¢*´ñ¶ýŸv,1—›+žþ§ˆ¶ùW§
+ÿ¶ùW+»Y·,1¸
*¶

¸
¶
¶*¶o™	
§f¶*´ñ¶
¶
¶
¶¶¶ùW*´ñ¶à §`6„
§ÿ]*´—›T,1—›
+þ¶ùW§
+ÿ¶ùW+»Y·,1¸
*¶

¸
¶
¶*¶ò¶
¶jø¶¶¶ùW„
§þ|§

*¶o™+
	¶ùW>6*´ñ¶ü¢‰*´ñ¶ýŸt,1—›+žþ§ˆ¶ùW§
+ÿ¶ùW+»Y·,1¸
*¶

¸
¶
¶*¶o™	
§f¶*´ñ¶
¶
¶
¶¶¶ùW„
„
§ÿq*´—›R,1—›
+þ¶ùW§
+ÿ¶ùW+»Y·,1¸
*¶

¸
¶
¶*¶ò¶
¶jø¶¶¶ùW+¶

°
OÞ723
567:9G:I;V<b=e;k@nAŒB§C¸DÑEäCèFïGþH

I
J
'L
.N
PO
yN
}Q
’G
˜S
¡T
ªU
´W
»Y
òA
ø^
û_`
bcd'e/fChJjkk”j˜m›c¡oªp²q¼sÃuùz
PpL
˜
<ò¦
Ù
<ï

»
<q
‡
˜
<I
¯
Ú
<n
Š
Û
<’
Ù
<í
»
<þ
Q
Rì
ž
Ü
ä
Ý
A
b
j(
ÿA
Ô
Þ
Ã

úý

P
ÿ
Ô
Þ
Ã




U
Ãÿ

Ô
Þ
Ã



Ã
ÿ+
Ô
Þ
Ã



Þ
 ÿ
Ô
Þ
Ã



Þ
 
jý

k
Þÿ

Ô
Þ
Ã





Þ
jÿ-
Ô
Þ
Ã





Þ
 ÿ

Ô
Þ
Ã





Þ
 
jP
ÿ
Ô
Þ
Ã






úú6øý

j
Þÿ

Ô
Þ
Ã


Þ
jÿ,
Ô
Þ
Ã


Þ
 ÿ

Ô
Þ
Ã


Þ
 
júú5

ß
V

N.


°
O
ƒ
P

Q
R	
à
f

N:
»
Y·

*¸
±
O
Œ
P

á
h
â
M

N
#ß½
Y»
Y²±¶V
·
SY»
Y²¶V
·
SY»
Y²
¶V
·
SY»
Y²
¶V
·
SY»
Y²
¶V
·
SY»
Y²
¶V
·
SY»
Y²³¶V
·
SY»
Y²²¶V
·
SY»
Y²
¶V
·
SY	»
Y²
¶V
 ·
SY
»
Y²
!¶V
"·
S³S±

O2µ!¶4·G¸Z¹mº»•¼©½½¾Ñ¿Þµ
ã
ä
ýÿ@@š˜Z@PK

Îl‹K	íAMETA-INF/þÊPK

Íl‹KøÈ÷wüü¤+META-INF/MANIFEST.MFPK

Îl‹KíAY
main/PK

Îl‹K
íA|
main/java/PK

Îl‹KíA¤
weka/PK

Îl‹KíAÇ
weka/classifiers/PK

Îl‹KíAö
weka/classifiers/functions/PK

Îl‹K åþ5T5T*¤/weka/classifiers/functions/LibLINEAR.classPKû
¬V