• No results found

Prediction instrument development for complex domains

N/A
N/A
Protected

Academic year: 2021

Share "Prediction instrument development for complex domains"

Copied!
226
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

(2) 35(',&7,21 ,167580(17 '(9(/230(17 )25 &203/(; '20$,16. 6MRHUG YDQ GHU 6SRHO.

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± &HQWUH IRU 7HOHPDWLFV DQG ,QIRUPDWLRQ 7HFKQRORJ\ 32 %R[  $( (QVFKHGH 7KH 1HWKHUODQGV 7KLV UHVHDUFK ZDV FRQGXFWHG LQ SDUW ZLWKLQ WKH &0RUH SURMHFW VXSSRUWHG E\ 'LQDORJ WKH 'XWFK ,Q VWLWXWH IRU $GYDQFHG /RJLVWLFV. ,6%1 ,661 '2, 85/ 3ULQWHU. ±±±±  &7,7 3K' 7KHVLV 6HULHV 1R ±

(4)  ?iiT,ff/tX/QBXQ`;fRyXjNNyfRXNd3Nyje89Rd9N ,SVNDPS 3ULQWLQJ. &RS\ULJKW ‹  6MRHUG YDQ GHU 6SRHO.

(5) 35(',&7,21 ,167580(17 '(9(/230(17 )25 &203/(; '20$,16. 352()6&+5,)7. WHU YHUNULMJLQJ YDQ GH JUDDG YDQ GRFWRU DDQ GH 8QLYHUVLWHLW 7ZHQWH RS JH]DJ YDQ GH UHFWRU PDJQLILFXV SURIGU + %ULQNVPD YROJHQV EHVOXLW YDQ KHW &ROOHJH YRRU 3URPRWLHV LQ KHW RSHQEDDU WH YHUGHGLJHQ RS ZRHQVGDJ  VHSWHPEHU RP  XXU. GRRU 6MRHUG -XUULDQ YDQ GHU 6SRHO JHERUHQ RS  QRYHPEHU  WH $OPHOR 1HGHUODQG.

(6) 7KLV GLVVHUWDWLRQ KDV EHHQ DSSURYHG E\ 3URIGU - YDQ +LOOHJHUVEHUJ SURPRWRU 'U & $PULW FRSURPRWRU. &RS\ULJKW ‹  6MRHUG YDQ GHU 6SRHO ,6%1 ±±±±.

(7)

(8) &RQWHQWV 3DUW , å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

(9) 3DUW ,, å &DVH VWXGLHV  6RIWH[FOXVLYH GRPDLQ DQDO\VLV IRU D VLPSOH GRPDLQ  &DVH GHVFULSWLRQ DVVXPSWLRQV DQG VLPXODWLRQ FRQVWUXFWLRQ  6LPXODWHG WHUPLQDO                    6WDJH , $VVXPSWLRQV                       *RDO GHĆQLWLRQ                      'DWD FRQVWUDLQW                      6WDJH ,, 3UHGLFWLYH PRGHOLQJ                   'DWD VHOHFWLRQ VWUDWHJLHV                 ([SORUDWRU\ GDWD DQDO\VLV                 7HFKQLTXH VHOHFWLRQ                    0RGHO WUDLQLQJ                      'LVFXVVLRQ                             &RQFOXVLRQ                            6RIWH[FOXVLYH GRPDLQ DQDO\VLV IRU D FRPSOH[ GRPDLQ  &DVH GHVFULSWLRQ                    6WDJH , $VVXPSWLRQV                  *RDO GHĆQLWLRQ                 /LWHUDWXUH 5HYLHZ                6XUYH\ UHVXOWV                  'DWD FRQVWUDLQW                 6WDJH ,, 3UHGLFWLYH PRGHOLQJ              'DWD VHOHFWLRQ FOHDQLQJ VWUDWHJLHV      5HGXFWLRQ E\ GDWD FRQVWUDLQW          ([SORUDWRU\ GDWD DQDO\VLV            7HFKQLTXH VHOHFWLRQ               (YDOXDWLRQ YDOLGDWLRQ PRGHO VHOHFWLRQ    6WDJH ,,, 0RGHO FRQYHUJHQFH              'LVFXVVLRQ                        &RQFOXVLRQ                     .  6RIWLQFOXVLYH GRPDLQ DQDO\VLV IRU D FRPSOH[ GRPDLQ  3UHSDUDWLRQ                       3UREOHP LGHQWLĆFDWLRQ             ([SHUW VHOHFWLRQ                6WDJH ,                        .               .    .               .    .               .    .               .    .               .    .                           .    .                           .    .            .   . .             .               .                 .    .     .

(10) .   .  +\SRWKHVLV GLYHUJHQFH                     +\SRWKHVLV FRQYHUJHQFH                    &RQVWUDLQW GHĆQLWLRQ                     6WDJH ,,                                 'DWD VHOHFWLRQ FOHDQLQJ VWUDWHJLHV              5HGXFWLRQ E\ GDWD GRPDLQ FRQVWUDLQWV            ([SORUDWRU\ GDWD DQDO\VLV                    7HFKQLTXH SDUDPHWHU VHOHFWLRQ               0RGHO WUDLQLQJ                         5HGXFWLRQ E\ LQWHUHVWLQJQHVV GHSOR\PHQW GRPDLQ FRQ VWUDLQWV                            6WDJH ,,,                                'LVFXVVLRQ                               &RQFOXVLRQ                             .             . 3DUW ,,, å &RQFOXVLRQ  'LVFXVVLRQ   &RQWULEXWLRQ                               &RQFOXVLRQ  $QVZHUV WR WKH UHVHDUFK TXHVWLRQV   /LPLWDWLRQV               )XWXUH ZRUN               9DOLGDWLRQ            'LUHFWLRQV          .     .     .     .     .     .     .     .     .     .     .     .     .     .     .     .     .      . $SSHQGLFHV $ 6XPPDU\. . % 6DPHQYDWWLQJ. . & 3XEOLFDWLRQV. . ' %LEOLRJUDSK\. .

(11) ( 'HWDLOV RI WKH GRPDLQ DQDO\VLV UHYLHZ. . ) 'HWDLOV RI WKH DUULYDO WLPH UHYLHZ. . * 'LVWULEXWLRQ FHQWHU VXUYH\. . + 'LVWULEXWLRQ FHQWHU LQWHU YDULDEOH FRUUHODWLRQV. . ,. 'HWDLOV RI WKH WXUQDURXQG WLPH UHYLHZ. . -. 'RPDLQ FRPSOH[LW\ DVVHVVPHQW  - &DOFXODWLQJ GRPDLQ FRPSOH[LW\                    . . /LVW RI MRXUQDOV. . / &RQFLVH PHWKRG KDQGERRN. . 0 'HVLJQ OD\RXW  0 $ERXW WKH FRYHU                            . .

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¶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µVHFUHWV¶ 7KLV WXUQHG LQWR D ELW RI D FKLFNHQ DQG HJJ SUREOHP FRQYLQFLQJ WKH FRPSDQLHV WR VXUUHQGHU GDWD UHTXLUHG D FOHDU LGHD RI ZKDW , FRXOG RIIHU WKHP DQG JHWWLQJ D FOHDU LGHD RI ZKDW , FRXOG RIIHU UHTXLUHG VRPH WHVW GDWD (YHQWXDOO\ WKLV VWDOHPDWH EURNH DV , IRXQG D GLVWULEXWLRQ FHQWHU FRPSDQ\ ZLOOLQJ WR FROODERUDWH ,Q WKH PRQWKV EHIRUH , KDG EHHQ WHOOLQJ P\VHOI ³LI RQO\ , KDG VRPH UHDOZRUOG GDWD´ $IWHU ZRUNLQJ RQ WKLV UHDOZRUOG GDWD DQG VHHLQJ LWV LQKHUHQW PHVVLQHVV , IRXQG P\VHOI WKLQNLQJ ³LI RQO\ , KDG FOHDQ VLPXODWHG GDWD´ RU ³LI RQO\ , GLGQ¶W KDYH WR ZRUN ZLWK GDWD´ +RZHYHU WKLV ILUVW FDVH WXUQHG RXW WR EH WKH NH\ WR WKLV WKHVLV ,W VKRZHG WKH GLIILFXOW\ RI ZRUNLQJ ZLWK GDWD IURP UHDOZRUOG VHWWLQJV DQG WKH GLIILFXOW\ RI GHULYLQJ SUHGLFWLYH PRGHOV IURP WKRVH VHWWLQJV 6ORZO\ EXW VXUHO\ D PHWKRG .

(13) HPHUJHG IRU GHDOLQJ ZLWK WKHVH UHDOZRUOG SUREOHPV LQ D VWUXFWXUHG ZD\ 6R WKH ELJ SUREOHP RI QRW EHLQJ DEOH WR JHW JRRG

(14) GDWD EXW KDYLQJ WR FUHDWH SUH GLFWLYH PRGHOV ZLWK LW DQ\ZD\ WXUQHG RXW WR EH WKH ULJKW GLUHFWLRQ IRU P\ WKH VLV 7KDW ZDV RQH RI WKH PDLQ OHDUQLQJ SRLQWV RI 3K' SURMHFW PDNH WKH EHVW RXW RI ZKDW \RX KDYH LQVWHDG RI WU\LQJ WR JR IRU SHUIHFWLRQ ,W ZDV YHU\ YDOXDEOH WR OHDUQ WR GHDO ZLWK SRRU TXDOLW\ GDWD DQG XQSUHGLFWDEOH RXWFRPHV SHUKDSV PRUH YDOXDEOH WKDQ ZRUNLQJ ZLWK FOHDQ DQG WLG\ GDWD ZRXOG HYHU KDYH EHHQ $OO LQ DOO WKH 3K' SURFHVV ZDV TXLWH D µULGH¶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

(15)  DQG WKDQNV IRU WKH HYHU PRUH DPELWLRXV MRXUQDO VXJJHVWLRQV , DOVR ZDQW WR WKDQN P\ IHOORZ 3K' VWXGHQWV 0\ RIILFH PDWHV $EKLVKWD $QGUHM *UpDQQH )ORRU /XFDV 5RELQ 6LPRQ DQG :RXWHU WKDQN \RX IRU WKH FRQ YHUVDWLRQV VRPH IXQ VRPH VHULRXV

(16)  IUHTXHQW FRIIHH EUHDNV DQG JRRG RIILFH DWPRVSKHUH 7KDQNV DOVR WR $UWXUR $UYLDQV\DK 'LVVD )DELDQ *XLGR 1LOV DQG :DKLG IRU WKH OXQFK ZDONV DQG OXQFK GLVFXVVLRQV DERXW HYHU\WKLQJ DQG QRWKLQJ 0\ WKDQNV JR RXW WR P\ SDUHQWV IRU WKHLU HPRWLRQDO VXSSRUW WKURXJKRXW WKH 3K' SURMHFW DQG IRU WKHLU ILQDQFLDO VXSSRUW WKURXJKRXW P\ SXUVXLW RI D %DFKHORU¶V DQG 0DVWHU¶V GHJUHH :LWKRXW WKHLU VXSSRUW , TXLWH OLWHUDOO\ FRXOG QRW KDYH GRQH D 3K' .

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

(18) .

(19)  ,QWURGXFWLRQ 7KH GHYHORSPHQW RI SUHGLFWLRQ LQVWUXPHQWV GHSHQGV VWURQJO\ RQ WKH GRPDLQ IRU ZKLFK WKH\ PDNH SUHGLFWLRQV 7KH PRUH FRPSOH[ WKH GRPDLQ WKH KDUGHU LW LV WR PDNH JRRG SUHGLFWLRQV DV D FRPSOH[ GRPDLQ KDV PDQ\ IDFWRUV WKDW LQIOX HQFH WKH SUHGLFWLRQ RXWFRPH   ZKLFK PHDQV WKHUH DUH PRUH ³XQFRQWUROOHG IDFWRUV >DQG@ QRLVH´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µZKDW QHHGV SUHGLFWLQJ¶ ZKLFK DOORZV WKH GHYHORSHU WR GHWHUPLQH ZKLFK YDULDEOHV VKRXOG EH LQFOXGHG LQ SUHGLFWLYH PRGHOV 7KLUG XQGHUVWDQGLQJ ZKLFK SUHGLFWLYH PRGHOV DUH XVDEOH DQG LQWHUHVW LQJ ZLWKLQ WKH GRPDLQ LQ WHUPV RI SUHGLFWLYH SRZHU LQWHJUDWLRQ ZLWK H[ LVWLQJ LQIUDVWUXFWXUH DQG LQWHJUDWLRQ ZLWK EXVLQHVV UXOHV DQG SURFHVVHV 7KHUH LV D PXOWLWXGH RI PHWKRGV IRU SUHGLFWLRQ LQVWUXPHQW GHYHORSPHQW VXFK DV 6KPXHOL DQG .RSSLXV¶  PHWKRG IRU SUHGLFWLYH DQDO\WLFV DQG WKH &URVV ,QGXVWU\ 6WDQGDUG 3URFHVV IRU 'DWD 0LQLQJ &5,63'0

(20)   WKDW DFNQRZOHGJH WKH LPSRUWDQFH RI XQGHUVWDQGLQJ WKH FRQWH[W D SUHGLFWLRQ LQVWUXPHQW LV GHYHO RSHG IRU &5,63'0 IHDWXUHV %XVLQHVV 8QGHUVWDQGLQJ DQG 'DWD 8QGHUVWDQG LQJ VWHSV WKDW KDYH D VLPLODU SXUSRVH WR WKDW GHVFULEHG DERYH 7KH VDPH DS SOLHV WR 6KPXHOL DQG .RSSLXV¶ PHWKRG ZKHUH WKDW SXUSRVH LV DFKLHYHG E\ WKH .

(21) i 5HVHDUFK TXHVWLRQV. *RDO 'HILQLWLRQ 'DWD &ROOHFWLRQ DQG 6WXG\ 'HVLJQ ([SORUDWRU\ 'DWD $QDO\ VLV DQG 9DULDEOH 6HOHFWLRQ VWHSV :KDW QHLWKHU PHWKRG VWDWHV KRZHYHU LV H[ DFWO\ ZKDW NQRZOHGJH VKRXOG EH JDWKHUHG RU KRZ WKDW LV WR EH JDWKHUHG WKH\ PHUHO\ VWDWH WKDW GRPDLQ XQGHUVWDQGLQJ VKRXOG EH DGGUHVVHG  S S  Så  7KH UHPDLQGHU RI WKLV LQWURGXFWLRQ LV RUJDQL]HG DV IROORZV )LUVW ZH SUHVHQW WKH WKUHH PDLQ UHVHDUFK TXHVWLRQV WKDW DUH DGGUHVVHG WKURXJKRXW WKLV WKHVLV 1H[W ZH SUHVHQW D WKHRUHWLFDO EDFNJURXQG RI WKH FRQFHSW DQG LPSRUWDQFH RI GRPDLQ DQDO\VLV :KDW IROORZV LV D OLWHUDWXUH UHYLHZ RI WKH µVWDWHRIWKHDUW¶ LQ WHUPV RI GRPDLQ DQDO\VLV IRU SUHGLFWLRQ LQVWUXPHQW GHYHORSPHQW WR FODULI\ WKH JDS WKDW WKLV WKHVLV DGGUHVVHV 1H[W ZH SUHVHQW WKHRU\ IRU WKH W\SH RI LQ WHOOLJHQFH WKDW VKRXOG EH JDWKHUHG LQ GRPDLQ DQDO\VLV 7KLV FKDSWHU HQGV ZLWK DQ RYHUYLHZ RI WKH WHUPLQRORJ\ XVHG WKURXJKRXW WKH WKHVLV DQG DQ RYHUYLHZ RI WKH VWUXFWXUH RI WKH WKHVLV. i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µIURP WKH JURXQG XS¶ VR WKDW SUHGLF WLYH SRZHU LV PD[LPL]HG 7R DQVZHU WKLV TXHVWLRQ ZH PXVW XQGHUVWDQG KRZ GRPDLQ DQDO\VLV VKRXOG EH SHUIRUPHG ZKDW PHWKRGV IRU GRPDLQ DQDO\VLV ILW ZLWK ZKLFK GRPDLQV DQG ZKDW DVSHFWV RI GRPDLQ DQDO\VLV SRVLWLYHO\ DIIHFW SUHGLFWLYH SHUIRU PDQFH 7KLV TXHVWLRQ LV DGGUHVVHG LQ WKH WKHRUHWLFDO EDFNJURXQG SUH VHQWHG LQ WKLV FKDSWHU DQG LQ WKH FDVH VWXGLHV SUHVHQWHG LQ &KDSWHUV   DQG  :KDW LV WKH UHODWLRQ EHWZHHQ GRPDLQ W\SH GRPDLQ DQDO\VLV W\SH DQG SUH GLFWLYH SRZHU" +RZ ZHOO D SUHGLFWLRQ LQVWUXPHQW FDQ PDNH SUHGLFWLRQV GHSHQGV DPRQJVW RWKHU WKLQJV RQ KRZ GLIILFXOW WKH SUHGLFWLRQ SUREOHP .

(22) ,QWURGXFWLRQ. LV WKDW LW LV WU\LQJ WR VROYH 7KH SUHGLFWLRQ SUREOHP LV V\QRQ\PRXV ZLWK WKH GRPDLQ WKH SUHGLFWLRQ LQVWUXPHQW RSHUDWHV LQ VHH WKH WHUPLQRORJ\ LQ 6HFWLRQ 

(23)  7KLV TXHVWLRQ LV DERXW ZKDW PDNHV D SUHGLFWLRQ SUREOHP D GLIILFXOWFRPSOH[ SUREOHP RU LQ RWKHU ZRUGV ZKDW PDNHV D GRPDLQ D GLIILFXOW GRPDLQ WR PDNH SUHGLFWLRQV DERXW 7R DGGUHVV WKDW TXHVWLRQ LW LV QHFHVVDU\ WR FDWHJRUL]H GRPDLQV LQWR µVLPSOHWRSUHGLFW¶ DQG µGLIILFXOW WRSUHGLFW¶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

(24) i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Ć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

(25) ,QWURGXFWLRQ. A B C D 1 2 3. 6\VWHP 'RPDLQ. )LOWHU. 'DWD. )LJXUH  7KH GRPDLQ XVHG IRU D SUHGLFWLRQ LQVWUXPHQW LV UHSUHVHQWHG LQ GDWD WKURXJK D ILOWHU 7KLV ILOWHU FDQ EH DIIHFWHG E\ IDFWRUV IURP WKH GRPDLQ VXFK DV EHKDYLRUV JRDOV DQG SROLWLFV FKRLFH RI GRPDLQ DQDO\VLV W\SH 7R DQVZHU WKH TXHVWLRQ ZH LQYHVWL JDWH LI DQG KRZ WKH UHTXLUHPHQWV IRU D GRPDLQ DQDO\VLV PHWKRG IRU D VLPSOH SUREOHP GLIIHU IURP WKH UHTXLUHPHQWV IRU D GRPDLQ DQDO\ VLV PHWKRG IRU GLIILFXOWFRPSOH[ SUREOHPV. i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³D FRQFHSW RI D ZKROH ZKLFK KDV SURSHUWLHV DV D VLQJOHHQWLW\ VRFDOOHG µHPHUJHQW SURSHU WLHV´¶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

(26) i 7KHRUHWLFDO EDFNJURXQG. $EVWUDFW GRPDLQ $WWULEXWH  $WWULEXWH  $WWULEXWH . $WWULEXWH   $WWULEXWH  $WWULEXWH  $WWULEXWH  'RPDLQ LQVWDQFH. $WWULEXWH  $WWULEXWH   $WWULEXWH  $WWULEXWH   $WWULEXWH  'RPDLQ LQVWDQFH. )LJXUH  $EVWUDFW GRPDLQV DQG GRPDLQ LQVWDQFHV $ OLJKWQLQJ V\PERO 

(27) VLJQLILHV WKDW DQ DWWULEXWH LV DIIHFWHG E\ QRLVH +HUH WKH WZR GRPDLQ LQVWDQFHV DOVR ERWK KDYH H[WUD DWWULEXWHV WKDW DUH QRW VKDUHG EHWZHHQ LQVWDQFHV RU ZLWK WKH DE VWUDFW GRPDLQ. PHDQV WKDW UHOHYDQW DVSHFWV

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

(29)   +RZHYHU QHLWKHU PHWKRG FODU LILHV KRZ WKLV GRPDLQ XQGHUVWDQGLQJ VKRXOG EH GRQH .

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

(31) V\V WHPV HQJLQHHULQJ DW WKH HQG RI WKH V ,Q V\VWHPV HQJLQHHULQJ VRIW IDFWRUV VXFK DV EHKDYLRU SROLWLFV RU WUXVW DUH FRQVLGHUHG QRW UHOHYDQW DQG GRPDLQ DQDO\VLV FRQVLVWV RI WU\LQJ WR PRGHO WKH GRPDLQ PDWKHPDWLFDOO\ 2IWHQ WKDW PHDQV ILQGLQJ D SRSXODU SUREOHP WKDW LV VLPLODU DQG UHXVLQJ WKH DQDO\VLV PDGH WKHUH   ,Q VRIW V\VWHPV PHWKRGRORJ\ D GRPDLQ LV FRQVLGHUHG µD FRPSOH[ SUREOHP DWLFDO KXPDQ VLWXDWLRQ¶   DQG LW VWDWHV WKDW LQTXLU\ VKRXOG EH PDGH WR XQ GHUVWDQG HDFK DVSHFW RI WKDW GRPDLQ E\ WDONLQJ WR WKRVH LQYROYHG ,W LV QRW D UHSODFHPHQW RI V\VWHPV HQJLQHHULQJ EXW UDWKHU DQ H[WHQVLRQ DV &KHFNODQG ZULWHV ³>6RIW V\VWHPV PHWKRGRORJ\@ GHULYH>V@ ERWK IURP LQWHUSUHWLQJ WKH IDFWV DQG ORJLF RI D VLWXDWLRQ DQG IURP HQJDJLQJ ZLWK WKH P\WKV DQG PHDQLQJV WKURXJK ZKLFK RWKHU PDQDJHUV PDNH VHQVH RI LW´  S  )RUUHVWHU¶V V\VWHP G\QDPLFV DJUHHV ZLWK &KHFNODQG¶V YLHZV RI UHMHFWLQJ WKH XVH RI PDWKHPDWLFDO VLPSOLILFDWLRQ IRU DQDO\]LQJ FRPSOH[ GRPDLQV   )RU UHVWHU EHOLHYHV WKDW WKLV W\SH RI VRIWH[FOXVLYH RU µVWDWLF¶ DSSURDFK FDQQRW FDS WXUH WKH FRPSOH[LWLHV RI GRPDLQV LQ WKH UHDO ZRUOG +RZHYHU )RUUHVWHU VWDWHV WKDW VRIW V\VWHP PHWKRGRORJ\¶V TXDOLWDWLYH PHWKRGV RI DQDO\VLV ODFN ULJRU DV WKH\ GHSHQG RQ WKH UHVHDUFKHU¶V LQWXLWLRQV   6\VWHP G\QDPLFV DLPV WR UH VROYH WKLV E\ LWV µGLVFLSOLQLQJ IUDPHZRUN¶  S RI PRGHO IRUPXOL]DWLRQ GRPDLQ GHVFULSWLRQ DQG PRGHO WHVWLQJGRPDLQ VLPXODWLRQ %RWK )RUUHVWHU DQG &KHFNODQG VWDWH WKDW DSSURDFKHV WKDW UHXVH H[LVWLQJ GRPDLQ DQDO\VHV DUH QRW DSSOLFDEOH IRU FRPSOH[ GRPDLQV    EXW VXJJHVW WKDW WKLV LV QRW SURE OHPDWLF IRU VLPSOH GRPDLQV )RU H[DPSOH FRQVLGHU SUHGLFWLQJ WUXFN WUDYHO WLPH 7KH GRPDLQ IRU WKLV FDVH LV DQ LQVWDQFH RI WKH µWUDYHO WLPH¶ GRPDLQ ZKLFK LV WKRURXJKO\ FRYHUHG LQ OLWHUDWXUH $ VRIWH[FOXVLYH DSSURDFK ZRXOG EH WR WDNH WKLV OLWHUDWXUH DQG XVH LW DV D EDVLV IRU YDULDEOH VHOHFWLRQ DQG GDWD FOHDQLQJ 7KH PRVW LPSRUWDQW YDULDEOHV ZRXOG WKHQ EH ZHDWKHU DQG WUDIILF   ,I WKHUH DUH LQ IDFW QR RWKHU .

(32) i 7KHRUHWLFDO EDFNJURXQG. IDFWRUV WKDW FRXOG DIIHFW WKH RXWFRPH WKHQ WKLV VRIWH[FOXVLYH DQDO\VLV FRXOG JLYH JRRG UHVXOWV EXW DV SUHYLRXV ZRUN VKRZV   LI WKHUH DUH EHKDYLRUDO LQ IOXHQFHV VXFK DV WKDW WKH GULYHUV KDYH QR LQWHQWLRQ WR EH SXQFWXDO

(33)  SUHGLFWLYH SRZHU LV VHYHUHO\ UHGXFHG 7KLV VKRZV WKDW D GRPDLQ DQDO\VLV IURP DQ DEVWUDFW GRPDLQ FDQQRW EH WUDQVODWHG RQHWRRQH WR DQ LQVWDQFH RI WKDW GRPDLQ WKDW LV VXEMHFW WR EHKDYLRU DQG RWKHU µVRIW¶ IDFWRUV )LJXUH  IXUWKHU LOOXVWUDWHV WKLV SRLQW ,Q WKH ILJXUH WKHUH DUH WZR LQ VWDQFHV RI DQ DEVWUDFW GRPDLQ 7KH DEVWUDFW GRPDLQ KDV D PLQLPDO VHW RI DW WULEXWHV WKDW HYHU\ RQH RI LWV LQVWDQFHV DOVR KDV +RZHYHU DV LQVWDQFHV UHS UHVHQW FRPSOH[ GRPDLQV WKH\ DUH DIIHFWHG E\ RWKHU IDFWRUV WKDW RQO\ DSSO\ WR WKDW VSHFLILF GRPDLQ LQVWDQFH VXFK DV FXOWXUH SROLWLFV DQG EHKDYLRUDO LQIOX HQFHV )XUWKHUPRUH LQ LQVWDQFHV RI WKH GRPDLQ WKH VDPH DWWULEXWHV PLJKW EH GLIIHUHQWO\ DIIHFWHG E\ µQRLVH¶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i 6RIW KDUG IDFWRUV. )DFWRUV H[LVW RQ D VSHFWUXP IURP KDUG WR VRIW 7KH OHVV LQIOXHQFHG D IDFWRU LV E\ WKH SHRSOH SROLWLFV DQG EHKDYLRU LQ WKH GRPDLQ WKH PRUH KDUG LW LV 7KHVH IDFWRUV WKHPVHOYHV LQIOXHQFH ZKDW LV WR EH SUHGLFWHG RU LQIOXHQFH WKH TXDOLW\ RI GDWD RU LQIRUPDWLRQ WKDW LV XVHG WR PDNH SUHGLFWLRQV %HORZ ZH ZLOO JLYH D GHILQLWLRQ RI D IDFWRU DQG ZKDW PDNHV LW µKDUG¶ RU µVRIW¶ :H XVH WKH IROORZLQJ GHILQLWLRQ RI D IDFWRU D SKHQRPHQRQ RU FRQVWUXFW IURP WKH GRPDLQ WKDW DIIHFWV ZKDW LV WR EH SUHGLFWHG $ IDFWRU FRXOG EH D YDUL DEOH RU D W\SH RI QRLVH DIIHFWLQJ WKH UHODWLRQ EHWZHHQ YDULDEOHV RU EHWZHHQ D YDULDEOH DQG ZKDW LV WR EH SUHGLFWHG VXFK DV WKH W\SHV GLVFXVVHG LQ   7KH .

(34) ,QWURGXFWLRQ. ODWWHU LV V\QRQ\PRXV ZLWK LQIRUPDWLRQ TXDOLW\ VXFK DV GLVFXVVHG LQ    ,I VXFK D IDFWRU LV GLUHFWO\ RU LQGLUHFWO\ LQIOXHQFHG E\ EHKDYLRU SROLWLFV DQG FXO WXUH LW LV FRQVLGHUHG D VRIW IDFWRU 3ROLWLFV FDQ EH GHILQHG DV SRZHU LQ DFWLRQ 3RZHU LV WKH DELOLW\ WR JHW VRPH RQH WR GR WKH\ ZRXOG RWKHUZLVH QRW GR RU WKH DELOLW\ WR SUHYHQW WKHP IURP GRLQJ VRPHWKLQJ WKH\ ZDQW WR GR   3ROLWLFV DUH WDNHQ WR EHFRPH UHOHYDQW LQ FLUFXPVWDQFHV ZKHUH GHFLVLRQV DUH ³VXEMHFW WR QHJRWLDWLRQ FRQWHVWDEOH >RU@ PRXOGDEOH´  S  DQGRU ZKHQ D GRPDLQ LQYROYHV LQGLYLGXDOV RU JURXSV ZLWK GLIIHUHQW PRWLYDWLRQV ZKHQ WKHUH DUH GLIIHUHQW YLHZV RQ ³WKH VWDWH RI WKH ZRUOG DQG ZKDW WKH IXWXUH KROGV´  S  RU ZKHQ FRQGLWLRQV LQ WKH GRPDLQ DUH XQFHUWDLQ 'DYHQSRUW HW DO  GHVFULEH LQIRUPDWLRQ SROLWLFV DV DQ H[DPSOH RI D VRIW IDFWRU WKDW DIIHFWV LQIRUPDWLRQ TXDOLW\ 7KHVH LQIRUPDWLRQ SROLWLFV DUH DERXW KRZ SDUWV RI WKH RUJDQL]DWLRQ WKDW DOO KDYH WKHLU RZQ ³FXOWXUH OHDGHUV DQG HYHQ YRFDEXODU\´  S VKDUH RU ZLWKKROG LQIRUPDWLRQ 7UXVW LV DQRWKHU VRIW IDFWRU WKDW DIIHFWV LQIRUPDWLRQ TXDOLW\   &KHSDLWLV IRXQG D 6RYLHWHUD SUHMXGLFH DJDLQVW GDWD VKDULQJ ³URRWHG LQ WKH EHOLHI WKDW SRZHU DQG VHFXULW\ DUH HURGHG LI GDWD LV VKDUHG´  S  WKDW ZDV VWUHQJWKHQHG E\ D IHDU WKDW UH YHDOLQJ DVVHWV WKURXJK VKDULQJ GDWD

(35) WR WKH JRYHUQPHQW FRXOG UHVXOW LQ KLJKHU WD[HV $V PHQWLRQHG DERYH WKH GLVWLFWLRQ KDUG±VRIW LV QRW ELQDU\ EXW UDWKHU D VSHFWUXP :HDWKHU FRXOG EH D KDUG IDFWRU LQIOXHQFLQJ DUULYDO WLPH LW LV QRW VXEMHFW WR EHKDYLRU RU SROLWLFV 2Q WKH RWKHU HQG RI WKH VSHFWUXP µPRWLYDWLRQ¶ LV D VRIW IDFWRU LW GLUHFWO\ UHIOHFWV D EHKDYLRUDO DVSHFW ,Q EHWZHHQ DUH FRPEL QDWLRQV LH IDFWRUV WKDW DUH QHLWKHU IUHH IURP EHKDYLRUDO LQIOXHQFH QRU D GLUHFW UHIOHFWLRQ RI EHKDYLRU &RQJHVWLRQ FRXOG EH DQ H[DPSOH DV LW LV QRW QHFHVVDU LO\ D GLUHFW FRQVHTXHQFH RI EHKDYLRU EXW EHKDYLRU FRXOG FKDQJH LWV HIIHFW RQ DUULYDO WLPH VRPH GULYHUV DUH PRUH DGHSW DW DYRLGLQJ WUDIILF MDPV $ VRIW IDFWRU LV D IDFWRU WKDW LV XQLTXH WR RQH GRPDLQ WKH\ UHSUHVHQW ³XQLTXH LGLRV\QFUDFLHV ZKLFK FKDUDFWHULVH KXPDQ VLWXDWLRQV´  S  $OWKRXJK WKH VDPH IDFWRU FRXOG EH DW SOD\ LQ WZR GRPDLQV LWV HIIHFW ZLOO EH VOLJKWO\

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

(37)  DQG SUREDEO\ QRW HYHQ WKH VDPH IRU HDFK RUJDQL]DWLRQ WKH DXWKRU IRXQG WKDW WKHUH ZDV PRUH WUXVW LQ VRPH UHJLRQV WKDQ LQ RWKHUV .

(38) i 7KHRUHWLFDO EDFNJURXQG. 6LPLODUO\ 'DYHQSRUW HW DO VWDWH WKDW WKH LQIOXHQFH RI LQIRUPDWLRQ SROLWLFV GHSHQGV SHU GRPDLQ 'DYHQSRUW HW DO GHVFULEH ILYH GLIIHUHQW PRGHOV RI LQIRU PDWLRQ SROLWLFV WKDW DOO KDYH D GLIIHUHQW LPSDFW RQ LQIRUPDWLRQ TXDOLW\ )RU H[DPSOH WKH µWHFKQRFUDWLFDO XWRSLDQLVP¶ PRGHO OHDGV WR ORZ LQIRUPDWLRQ TXDO LW\ ZKHUHDV µIHGHUDOLVP¶ LV WKRXJKW WR OHDG WR KLJKHU LQIRUPDWLRQ TXDOLW\   7KHUHIRUH LW LV QHFHVVDU\ WR VWXG\ WKH GRPDLQ LQ GHWDLO WR DVVHVV ZKLFK PRGHO LV LQ XVH DQG LQ H[WHQVLRQ ZKDW VKRXOG EH H[SHFWHG LQ WHUPV RI LQIRUPDWLRQ TXDOLW\ 7KH VDPH FRXOG EH WKH FDVH IRU PRUH µKDUG¶ IDFWRUV VXFK DV WKH FRQJHVWLRQ GLVFXVVHG DERYH ,Q VRPH GRPDLQV GULYHUV FRXOG EH PRWLYDWHG WR DUULYH DV HDUO\ DV SRVVLEOH XUJLQJ WKHP WR DYRLG WUDIILF MDPV ZKLOVW LQ RWKHUV VHH &KDS WHU 

(39)  GULYHUV FRXOG ODFN WKDW LQFHQWLYH. i 'RPDLQ DQDO\VLV WHFKQLTXHV. :H UHIHU WR DQ DQDO\VLV WKDW LQFOXGHV VRIW IDFWRUV DV VRIWLQFOXVLYH DQG E\ H[WHQ VLRQ DQ DQDO\VLV WKDW H[FOXGHV VRIW IDFWRUV LH RQO\ WDNHV WKRVH LQWR DFFRXQW DW WKH µKDUG¶ HQG RI WKH VSHFWUXP LV VRIWH[FOXVLYH :H DOVR FRQVLGHU DQ DQDO\VLV WKDW GRHV QRW WDNH WKH GRPDLQVSHFLILF HIIHFWV RI D VRIW IDFWRU LQWR DFFRXQW DV VRIWH[FOXVLYH DQ DQDO\VLV WKDW GRHV QRW FRQVLGHU KRZ WKH EHKDYLRU FXOWXUH DQG SROLWLFV VSHFLILF WR D GRPDLQ DIIHFW ZKDW LV WR EH SUHGLFWHG FDQ EH FRQVLG HUHG WKH VDPH DV DQ DQDO\VLV WKDW GRHV QRW FRQVLGHU WKHVH IDFWRUV DW DOO )RU D VRIWH[FOXVLYH GRPDLQ DQDO\VLV D JRRG FDQGLGDWH WHFKQLTXH LV D OLW HUDWXUH UHYLHZ :KHQ DWWHQWLRQ LV SDLG WR FRPSOHWHQHVV WKLV ZLOO UHYHDO WKH UHOHYDQW DVSHFWV RI WKH JURXS RI GRPDLQV WKH GRPDLQ EHORQJV WR ,I WKH GRPDLQ LV VLPSOH WKHQ WKHVH DVSHFWV ZRXOG EH WUDQVODWDEOH IURP WKH DEVWUDFW GRPDLQ WR WKH VSHFLILF GRPDLQ DQG LW LV OHVV OLNHO\ WKDW UHOHYDQW DVSHFWV RI WKH VLPSOH GRPDLQ DUH PLVVHG (YHQ WKRXJK D VRIWLQFOXVLYH GRPDLQ DQDO\VLV DSSURDFK LV ZLGHO\ FRQVLG HUHG LPSRUWDQW IRU SUHGLFWLYH PRGHOLQJ     WKHUH DUH IHZ PHWKRGV WKDW PDNH GRPDLQ DQDO\VLV D FOHDU ULJRURXV DQG WUDQVSDUHQW SURFHVV 2IWHQ GRPDLQ XQGHUVWDQGLQJ LV DW EHVW DQ LPSOLFLW SDUW RI PRGHO GHYHORSPHQW 7KLV FRXOG HLWKHU PHDQ WKDW LW ZDV QRW QHFHVVDU\ EHFDXVH WKH GHYHORSHU LV DQ H[SHUW RQ WKH GRPDLQ RU WKDW LW ZDV QRW GRQH 5HJDUGOHVV LW VHHPV WKHUH LV D GHPDQG IRU WUDQVSDUHQW UHSHDWDEOH WHFKQLTXHV WKDW FDSWXUH UHOHYDQW NQRZOHGJH IURP WKH GRPDLQ IRU FRPSOH[ GRPDLQV .

(40) ,QWURGXFWLRQ i $ UHYLHZ RI GRPDLQ DQDO\VLV LQ FRPSOH[ GRPDLQV. ,Q WKH SUHYLRXV VHFWLRQ ZH KDYH GHVFULEHG WKH SXUSRVH RI GRPDLQ DQDO\VLV LQ SUHGLFWLRQ LQVWUXPHQW GHYHORSPHQW 7R ILQG KRZ GRPDLQ DQDO\VLV DQG HVSH FLDOO\ WR ILQG ZKDW W\SH RI DQDO\VLV VRIWLQFOXVLYH RU VRIWH[FOXVLYH

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¶ %DVNHW RI -RXUQDOV   WKH -RXUQDO &LWDWLRQ 5H SRUW RI WKH ,6,   WKH 8QLYHUVLW\ RI 6\GQH\ -RXUQDO 5DQNLQJV   6&,PDJR  DQG VHYHUDO SHHU UHYLHZHG VFLHQWRPHWULF SXEOLFDWLRQV RQ MRXUQDO UDQN LQJV        :H XVHG RQO\ MRXUQDOV DV ZH DUH LQWHUHVWHG LQ WKRVH SDSHUV WKDW DUH FRQVLGHUHG WKH EHVW LQ WKLV ILHOG DQG MRXUQDOV DUH WKH ILHOG¶V SUL PDU\ W\SH RI FRQWULEXWLRQ 8VLQJ WKH OLVW RI MRXUQDOV ZH H[HFXWHG WKH IROORZLQJ TXHU\ RQ 6FRSXV DQG :HE RI 6FLHQFH ³VXSSO\ FKDLQ´ 25 VFP 25 ORJLVWLFV

(42) $1' DQDO\ 25 ³ELJ GDWD´ 25 ³GDWD PLQLQJ´ 25 ³EXVLQHVV LQWHOOLJHQFH´ 25 ³GHFLVLRQ VXSSRUW´

(43) $1' SUHGLFW  UHVXOWLQJ LQ  SDSHUV 7KH DEVWUDFW LQFOXVLRQ FULWHULD DUH WKH IROORZLQJ WKH\ DUH UHODWHG WR RQH RU PRUH DVSHFWV RI WKH VXSSO\ FKDLQ DQG WKH\ PHQWLRQ VRPH W\SH RI SUHGLF WLRQ XVLQJ GDWD 7ZR UHYLHZHUV DQDO\]HG WKH DEVWUDFWV XVLQJ WKHVH FULWHULD DQG GHFLGHG IRU HDFK ZKHWKHU LW ZDV KLJKO\ UHOHYDQW &RPSDULQJ WKHVH GHFLVLRQV WKH UHYLHZHUV KDG VXIILFLHQW LQWHUUDWHUUHOLDELOLW\ &URQEDFK¶V α .ULS SHQGRUII¶V α

(44)  ,Q WKH ILQDO URXQG RI UHYLHZ ZH UHYLHZHG WKH IXOO WH[W RI WKH KLJKO\ UHOHYDQW SDSHUV 7KH LQFOXVLRQ FULWHULD DUH WKDW WKH SDSHU GLVFXVVHV SUHGLFWLRQ UHODWHG WR WKH VXSSO\ FKDLQ DQG WKDW HPSLULFDO GDWD ZDV XVHG WR PDNH WKRVH SUHGLFWLRQV $IWHU WKLV URXQG  SDSHUV UHPDLQHG .

(45) i 7KHRUHWLFDO EDFNJURXQG. $QDO\VLV W\SH. 1XPEHU RI SDSHUV. 6RIWH[FOXVLYH. 6RIWLQFOXVLYH.      &RPSOH[. 6LPSOH. 'RPDLQ W\SH. )LJXUH  5HVXOWV RI WKH OLWHUDWXUH UHYLHZ DQG GRPDLQ FRPSOH[LW\ DQDO\VLV 7DEOH ( RQ SDJH  VKRZV WKH SDSHUV ZH IRXQG IRU HDFK GRPDLQ DQG DQDO\VLV W\SH. 1H[W ZH LQYHVWLJDWHG WKH FRPSOH[LW\ RI WKH GRPDLQV LQ HDFK SDSHU XVLQJ DQ DVVHVVPHQW WRRO EDVHG RQ -DFNVRQ DQG .H\V¶ GHILQLWLRQ RI FRPSOH[LW\   7KH DVVHVVPHQW WRRO ZKLFK FRQVLVWV RI HLJKW TXHVWLRQV LV LQ $SSHQGL[ - LQ FOXGLQJ D IRUPXOD WR FDOFXODWH D FRPSOH[LW\ VFRUH 2QH DXWKRU DQVZHUHG WKH TXHVWLRQV IRU DOO  SDSHUV $QRWKHU UHVHDUFKHU GLG WKH VDPH IRU D UDQGRPO\ VHOHFWHG VXEVHW RI  SDSHUV WR UXOH RXW UDWHU ELDV 6XIILFLHQW LQWHUUDWHU UHOLDELO LW\ &URQEDFK¶V α  .ULSSHQGRUII¶V α 

(46) LQGLFDWHV WKDW WKH UHYLHZHUV ZHUH LQ DJUHHPHQW. 5HVXOWV )LJXUH  VKRZV WKH UHVXOWV RI WKH OLWHUDWXUH UHYLHZ $V PHQWLRQHG LQ WKH SUHYLRXV VHFWLRQ ZH UDQNHG HDFK SDSHU¶V GRPDLQ FRPSOH[LW\ 3DSHUV WKDW KDG D FRPSOH[LW\ VFRUH > 3.00 DUH FDWHJRUL]HG WR GHDO ZLWK D FRPSOH[ GR PDLQ SDSHUV ZLWK D VFRUH ≤ 3.00 DUH DERXW VLPSOH GRPDLQV DV WKLV LV KDOIZD\ EHWZHHQ VLPSOH DQG FRPSOH[ )LJXUH  VKRZV KRZ GRPDLQ FRPSOH[LW\ LV GLV WULEXWHG IRU DOO SDSHUV LQ WKH UHYLHZ 6LPLODUO\ ZH ORRNHG DW HDFK SDSHU¶V GR PDLQ DQDO\VLV DSSURDFK 3DSHUV WKDW XVH VRIW DQDO\VLV IRU YDULDEOH VHOHFWLRQ RU GDWD FOHDQLQJ DUH ODEHOHG VRIWLQFOXVLYH WKH RWKHUV DUH ODEHOHG VRIWH[FOXVLYH .

(47) 1XPEHU RI SDSHUV. ,QWURGXFWLRQ.           . . . . 'RPDLQ FRPSOH[LW\. . )LJXUH  &RPSOH[LW\ RI WKH SDSHUV LQ WKH OLWHUDWXUH UHYLHZ $OO SDSHUV ZLWK D FRPSOH[ LW\ RI  RU DERYH DUH FRQVLGHUHG FRPSOH[ SDSHUV ZKHUH  LV WKH ORZHVW SRVVLEOH FRPSOH[LW\ VFRUH DQG ILYH LV WKH KLJKHVW SRVVLEOH FRPSOH[LW\ VFRUH $V WKH ILJ XUH VKRZV WKHUH DUH QR SDSHUV ZLWK D FRPSOH[LW\ EHWZHHQ  DQG  6HH $SSHQGL[ - IRU WKH FDOFXODWLRQ RI WKH FRPSOH[LW\ VFRUH. .

(48) i 7KHRUHWLFDO EDFNJURXQG. 7KH PDMRULW\ RI SDSHUV   RI WKH WRWDO QXPEHU LQ WKH UHYLHZ

(49) GHDOV ZLWK D FRPSOH[ GRPDLQ 7KHVH SDSHUV GHDO ZLWK UHDOZRUOG FDVH VWXGLHV LQ WKH VXS SO\ FKDLQ UDQJLQJ IURP VFKHGXOLQJ SUREOHPV WR SUHGLFWLQJ PDFURHFRQRP\ ILJXUHV 7KH WZR SDSHUV WKDW GHDO ZLWK D VLPSOH GRPDLQ ERWK PDNH SUHGLFWLRQV DERXW D GRPDLQ PDGH XS RI RQO\ LQWHOOLJHQW DJHQWV ZKLFK DUH QRW VXEMHFW WR EHKDYLRUDO LQIOXHQFHV

(50)   

(51) RI WKH WRWDO SDSHUV DQG  

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

(53)  EXW LV DOVR XVHG WR PDNH SUHGLFWLRQV 7KHVH SUHGLFWLRQV XVXDOO\ DUH WKH DPRXQW RI HIILFLHQF\ LQFUHDVH LI WKH GHYHORSHG PDWKHPDWLFDO PRGHO LV XVHG EXW FRXOG DOVR EH IRUH FDVWV 7KH VL[ SDSHUV WKDW GLG LQFOXGH VRIW IDFWRUV PRVWO\ XVHG LQWHUYLHZV DV D PHDQV RI FDSWXULQJ GRPDLQ LQWHOOLJHQFH 5DEHOR HW DO XVHG )RUUHVWHU¶V V\VWHP G\QDPLFV DQG FUHDWHG D PRGHO RI D VXSSO\ FKDLQ EDVHG ERWK RQ H[LVWLQJ WKHRU\ DQG LQWHUYLHZV ZLWK WKRVH LQYROYHG   i 'RPDLQ LQWHOOLJHQFH. 7KH SDUDJUDSKV DERYH GLVFXVV ZK\ D VRIWLQFOXVLYH GRPDLQ DQDO\VLV ILWV ZLWK PRUH FRPSOH[ GRPDLQV 7KDW OHDYHV RSHQ ZKDW H[DFWO\ VKRXOG EH DQDO\]HG DQG KRZ WKDW DQDO\VLV VKRXOG EH PDGH LQ WKH VSHFLILF FRQWH[W RI SUHGLFWLRQ LQVWUXPHQW GHYHORSPHQW &DR HW DO  SURYLGH DQ DQVZHU WR ERWK WKH ZK\ DQG ZKDW TXHVWLRQ LQ WKHLU 'RPDLQ 'ULYHQ 'DWD 0LQLQJ '0

(54) PHWKRG $OWKRXJK WKHLU PHWKRG LV IRFXVHG RQ XQGHUVWDQGLQJ WKH GRPDLQ UDWKHU WKDQ PDNLQJ SUHGLFWLRQV DERXW WKH GRPDLQ LW FDQ EH DGDSWHG WR EH VXLWHG IRU SUHGLFWLRQ SXUSRVHV 7KH PDLQ DUJXPHQW SUHVHQWHG IRU '0 LV WKDW GDWD GRHV QRW H[LVW LQ D YDFXXP DQG WKDW XQGHUVWDQGLQJ WKH GRPDLQ RI WKH GDWD DQG JDWKHULQJ LQWHOOLJHQFH LV NH\ WR JHWWLQJ YDOXDEOH LQVLJKWV IURP GDWD   .

(55) ,QWURGXFWLRQ. 'RPDLQ ,QWHOOLJHQFH. 2 2UJDQL]DWLRQDO ,QWHOOLJHQFH ,Q 6RFLDO ,QWHOOLJHQFH +XPDQ ,QWHOOLJHQFH. 'DWD LQ LQWHOOLJHQFH 1HWZRUN LQWHOOLJHQFH. )LJXUH  7KH VL[ W\SHV RI GRPDLQ LQWHOOLJHQFH WKDW QHHG WR EH FDSWXUHG GXULQJ GRPDLQ DQDO\VLV   2UJDQL]DWLRQDO VRFLDO DQG KXPDQ LQWHOOLJHQFH KDYH FRQVLGHUDEOH RYHUODS MXVW DV GDWD DQG QHWZRUN LQWHOOLJHQFH ,Q WXUQ GRPDLQ LQWHOOLJHQFH HQFRPSDVVHV DQG H[WHQGV DOO ILYH LQWHOOLJHQFHV. .

(56) i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µPRRG¶ LV DQ LPSRUWDQW SUHGLFWRU EXW LI WKLV LV QRW DYDLODEOH LQ WKH GDWD DQ\ SUHGLFWLRQ LQVWUXPHQW WKDW UHOLHV RQ LW ZRXOG QRW EH XVDEOH 7KH LQWHUHVWLQJQHVV FRQVWUDLQW UHODWHV WR WKH SHUIRUPDQFH RI WKH SUHGLFWLRQ LQVWUXPHQW ,I WKH SHUIRUPDQFH VXFK DV DFFXUDF\ SUHFLVLRQ LV WRR ORZ RU HUURU RI WKH SUHGLFWLRQV LV WRR ORZ IRU DFFXUDF\ DQG SUHFLVLRQ

(57) RU WRR KLJK IRU WKH HUURU

(58)  WKH LQVWUXPHQW FDQQRW EH XVHG IRU GHFLVLRQ VXSSRUW .

(59) ,QWURGXFWLRQ. y y. cz. cz. cy. cy $FWLRQDEOH. cy cy. 1RW DFWLRQDEOH. cx. cx. x. z cx. cx. x. )LJXUH  3UHGLFWLRQ LQVWUXPHQW VSDFH DQG DFWLRQDEOH SUHGLFWLRQ LQVWUXPHQW VSDFH 7KH OHIW VLGH RI WKH LPDJH VKRZV RQH FRQVWUDLQW RQ WKH yD[LV ZLWK XSSHU ERXQG cy DQG ORZHU ERXQG cy  DQG VLPLODU IRU WKH x D[LV ZLWK ORZHU ERXQG cx DQG cx  7KH DUHD ZLWKLQ WKHVH ERXQGV LV WKH DFWLRQDEOH LQVWUXPHQW VSDFH FRORUHG LQ ZKLWH $V WKH ULJKW KDQG VLGH RI WKH LPDJH VKRZV WKLV WZRGLPHQVLRQDO VSDFH LV IXUWKHU ERXQG E\ FRQVWUDLQWV RQ WKH zD[LV )RU FRPSOHWHQHVV VDNH WKHUH DUH LQ IDFW IRXU FRQVWUDLQW GLPHQVLRQ EXW WKH wD[LV LV QRW SLFWXUHG IRU UHDVRQV RI FODULW\

(60) &RQVWUDLQW GLPHQVLRQV FDQ EH HLWKHU TXDQWLWDWLYH RU TXDOLWDWLYH. .

(61) i 7HUPLQRORJ\. 7KH GHSOR\PHQW FRQVWUDLQW UHODWHV WR WKH LQWHJUDWLRQ RI WKH SUHGLFWLRQ LQ VWUXPHQW ZLWK H[LVWLQJ LQIUDVWUXFWXUH ,QVWUXPHQWV WKDW DUH QRW VXSSRUWHG E\ H[LVWLQJ KDUGZDUH DQG VRIWZDUH FDQQRW EH LQWHJUDWHG ZLWK H[LVWLQJ GHFLVLRQ VXSSRUW V\VWHPV RU FDQQRW JLYH SUHGLFWLRQV LQ UHDO WLPH DUH GHSHQGLQJ RQ FRQWH[W

(62) QRW XVDEOH IRU GHFLVLRQ VXSSRUW )LQDOO\ WKH GRPDLQ FRQVWUDLQW UHODWHV WR RWKHU FRQVWUDLQWV IURP WKH GR PDLQ ZKHUH WKH LQVWUXPHQW ZLOO EH XVHG 7KHVH LQFOXGH UHJXODWRU\ PHDVXUHV LQWHJUDWLRQ ZLWK H[LVWLQJ EXVLQHVV SURFHVVHV DQG FRPSOLDQFH ZLWK EXVLQHVV UXOHV 7KH GRPDLQ FRQVWUDLQW DOVR GHWHUPLQHV ZKLFK SHUIRUPDQFH PHWULF LV WR EH XVHG DQG ZKDW W\SH RI SUHGLFWLRQV HJ FODVVLILFDWLRQ RU UHJUHVVLRQ

(63) WKH SUHGLFWLRQ LQVWUXPHQW VKRXOG PDNH &RQVLGHU WKH VSDFH RI DOO SRVVLEOH SUHGLFWLRQ LQVWUXPHQWV WKDW FRXOG EH GH YHORSHG EDVHG RQ WKH LQWHOOLJHQFH JDWKHUHG SUHGLFWLYH LQVWUXPHQW VSDFH )RU HDFK W\SH RI FRQVWUDLQW IRXU LQ WRWDO

(64) WKHUH LV RQH GLPHQVLRQ LQ WKLV VSDFH ZKLFK LV OLPLWHG E\ WKH FRQVWUDLQW &RPELQLQJ DOO IRXU FRQVWUDLQW GLPHQVLRQV UHVXOWV LQ WKH DFWLRQDEOH DUHD RI SUHGLFWLRQ LQVWUXPHQW VSDFH ,I D SUHGLFWLRQ LQVWUXPHQW IDOOV RXWVLGH WKH OLPLWV RI RQH RU PRUH FRQVWUDLQWV LW FDQ QR ORQJHU EH FRQVLGHUHG DFWLRQDEOH )LJXUH  VKRZV WKH SUHGLFWLRQ LQVWUXPHQW VSDFH LQ WKUHH GLPHQVLRQV DQG KRZ WKH FRQVWUDLQWV GHWHUPLQH ZKLFK DUHD RI WKDW VSDFH LV DFWLRQDEOH. i 7HUPLQRORJ\ ,Q WKLV VHFWLRQ ZH H[SODLQ WKH WHUPLQRORJ\ XVHG WKURXJKRXW WKH WKHVLV 7KH SXUSRVH RI WKH VHFWLRQ LV WR ERWK FODULI\ QHZ WHUPLQRORJ\ DQG WR H[SODLQ KRZ ZH XVH H[LVWLQJ WHUPLQRORJ\ 7KH WHUPV FRYHUHG LQ WKLV VHFWLRQ DUH SUHGLFWLYH PRGHOV SUHGLFWLRQ LQVWUXPHQWV GRPDLQ W\SHV DQG GRPDLQ DQDO\VLV W\SHV RI ZKLFK SUHGLFWLRQ LQVWUXPHQW LV WKH PDLQ QHZ WHUP LQWURGXFHG i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

(65) ,QWURGXFWLRQ. &ROXPQ UHSUHVHQWLQJ WKH JRDO YDULDEOH. 5RZ UHSUHVHQWLQJ LQVWDQFHV. 5RZ. $JH. *HQGHU. 6DYLQJV. 0RUWJDJH.     .     . 0 ) 0  . ¼  ¼  ¼   . )DOVH 7UXH 7UXH  . Q. . ). ¼ . )DOVH. &ROXPQV UHSUHVHQWLQJ YDULDEOHV )LJXUH  $Q H[DPSOH GDWDVHW ZKHUH WKH YDULDEOHV DUH UHSUHVHQWHG E\ FROXPQV DQG WKH LQVWDQFHV RU REVHUYDWLRQV

(66) DUH UHSUHVHQWHG E\ URZV ,Q WKLV H[DPSOH 0RUW JDJH LV WKH JRDO YDULDEOH 6R ZKHQ YDOXHV IRU HDFK RI WKH RWKHU YDULDEOHV DUH SXW LQWR WKH SUHGLFWLYH PRGHO LW RXWSXWV D YDOXH IRU WKH JRDO YDULDEOH. PHWKRG WDNHV D KLVWRULFDO GDWDVHW FRQVLVWLQJ RI ERWK WKH LQGHSHQGHQW YDULDEOHV DQG WKH YDULDEOH WKDW LV WR EH SUHGLFWHG LH WKH GHSHQGHQW YDULDEOH ,W WKHQ GH WHUPLQHV ZKLFK YDOXHV IRU WKH LQGHSHQGHQW YDULDEOHV EHVW PRVW FOHDQO\ DQG FRPSOHWHO\

(67) VSOLW WKH GHSHQGHQW YDULDEOH LQWR SDUWV VXFK DV LQWHUYDOV RU H[DFW YDOXHV

(68)  $QRWKHU JURXS RI WHFKQLTXHV LV FDOOHG HQVHPEOH WHFKQLTXHV RI ZKLFK $G D%RRVW   DQG 5DQGRP )RUHVWV  DUH H[DPSOHV 7KHVH WHFKQLTXHV WDNH D JURXS KHQFH HQVHPEOH

(69) RI RWKHU ZHDNHU

(70) SUHGLFWLYH PRGHOV DQG FRQVLGHUV WKH SUHGLFWLRQ RI HDFK RI WKRVH D YRWH 7KH µSRSXODU¶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

(71) i 7HUPLQRORJ\. YDULDEOH LH WKH YDULDEOH WKH LQVWUXPHQW ZLOO SUHGLFW 2I FRXUVH WKLV FROXPQ ZLOO RQO\ KDYH D YDOXH IRU KLVWRULFDOH[LVWLQJ URZV DV LW PXVW EH SUHGLFWHG IRU QHZ REVHUYDWLRQV LH URZV LQ WKH PDWUL[ UHSUHVHQWDWLRQ RI D GDWDVHW

(72)  7KH GDWD VHOHFWLRQ DQG UHILQHPHQW VWUDWHJ\ LV D GHILQLWLRQ RI ZKDW GDWD VKRXOG EH XVHG ZKLFK URZV DQG FROXPQV LQ WKH PDWUL[

(73) DQG KRZ WKLV GDWD LV WR EH SUR FHVVHG DQG FOHDQHG $ QXPEHU RI URZV GHWHUPLQHG E\ DQ DOJRULWKPVSHFLILF SDUDPHWHU

(74) RI WKH GDWD FDOOHG WKH WUDLQLQJ VHW ZKLFK LQFOXGHV WKH JRDO YDULDEOH FROXPQ LV IHG LQWR WKH DOJRULWKP WR FUHDWH D PRGHO ZKLFK FRXOG EH D QHXUDO QHWZRUN D GHFL VLRQ WUHH D UHJUHVVLRQ PRGHO HWF VHH 3UHGLFWLYH PRGHO

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i 'RPDLQ GRPDLQ FRPSOH[LW\ DQG GRPDLQ DQDO\VLV. 3UHGLFWLRQ LQVWUXPHQWV DUH FUHDWHG WR VROYH D SUHGLFWLRQ SUREOHP 7KH FRQWH[W RI WKLV SUREOHP ZKLFK LV DOVR WKH FRQWH[W RI WKH SUHGLFWLRQ LQVWUXPHQW XQGHU GHYHORSPHQW LV ³D FRQFHSW RI D ZKROH WKDW KDV µHPHUJHQW SURSHUWLHV´¶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¶ GHILQLWLRQ RI FRPSOH[LW\   ZKLFK LV EXLOW XS DV D FKHFNOLVW ZLWK  7KLV. ZD\ WKH PRGHO FDQ DOVR EH WHVWHG E\ IHHGLQJ LQ URZV IRU ZKLFK WKH JRDO YDULDEOH YDOXH LV NQRZQ 7KH SUHGLFWLRQV FDQ WKHQ EH FRPSDUHG WR WKH DFWXDO YDOXHV WR GHWHUPLQH WKH SHUIRUPDQFH RI WKH PRGHO. .

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µSUHGLFWLQJ DUULYDO WLPH¶ RU µSUHGLFWLQJ FXVWRPHU FKXUQ¶ ,Q RWKHU ZRUGV WKH\ DUH WZR LQVWDQFHV RI RQH DEVWUDFW GRPDLQ 7KH UHYLHZ ZRXOG UHVXOW LQ YDULDEOH VHOHFWLRQ DQG GDWD FOHDQLQJ VWUDWHJLHV WKDW KDYH EHHQ XVHG SUHYLRXVO\ ZKLFK DUH WKHQ UHSXUSRVHG IRU WKH QHZ GRPDLQ 7KLV DSSURDFK RI GRPDLQ DQDO\VLV ZRXOG QRW UHIOHFW GRPDLQVSHFLILF IDFWRUV DQG FRQGLWLRQV WKDW DIIHFW WKH JRDO YDULDEOH VXFK DV FXOWXUH SROLWLFV EHKDYLRU RU HWKLFV EXW ZRXOG OLNHO\ RQO\ LQFOXGH WKRVH YDULDEOHV WKDW DUH GLUHFWO\ TXDQ WLILDEOH $V WKLV W\SH RI GRPDLQ DQDO\VLV H[FOXGHV WKH VRIW GRPDLQVSHFLILF IDF WRUV RU WKH VRIW DVSHFWV WKDW LQIOXHQFH WKHVH TXDQWLILDEOH

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i 6WUXFWXUH 7KLV VHFWLRQ JLYHV DQ RXWOLQH RI WKH UHPDLQGHU RI WKH WKHVLV 7KLV FKDSWHU KDV GLVFXVVHG D WKHRUHWLFDO EDFNJURXQG RQ GRPDLQ W\SH GRPDLQ DQDO\VLV DQG WKHLU UHODWLRQ WR SUHGLFWLYH SRZHU ,Q VR GRLQJ WKH UHVHDUFK TXHVWLRQV WKDW LQ YROYH WKH ZK\ DQG ZKDW RI GRPDLQ DQDO\VLV DV SDUW RI SUHGLFWLRQ LQVWUXPHQW GHYHORSPHQW KDYH EHHQ DGUHVVHG IURP D WKHRUHWLFDO SHUVSHFWLYH .

(78) i 6WUXFWXUH. 7KH WKHRUHWLFDO EDFNJURXQG LQ WKLV FKDSWHU RXWOLQHV D K\SRWKHVLV RQ KRZ GRPDLQ DQDO\VLV DIIHFWV SUHGLFWLYH SRZHU 7KH K\SRWKHVLV LV WKDW D SUHGLFWLRQ LQVWUXPHQW GHYHORSHG IRU D FRPSOH[ GRPDLQ WKDW LV DIIHFWHG E\ EHKDYLRUDO LQIOXHQFHV RSHQ WR LWV HQYLURQPHQW SDUWLDOO\ REVHUYDEOH DQG KDV SDUWV ZLWK WKHLU RZQ JRDOV

(79) ZLWK VRIWLQFOXVLYH GRPDLQ DQDO\VLV WKDW WDNHV EHKDYLRUDO LQ IOXHQFHV SROLWLFV VXESDUW JRDOV DQG LWV HQYLURQPHQW LQWR DFFRXQW ZLOO KDYH KLJKHU SUHGLFWLYH SRZHU WKDQ D SUHGLFWLRQ LQVWUXPHQW GHYHORSHG IRU WKH VDPH GRPDLQ ZLWK VRIWH[FOXVLYH GRPDLQ DQDO\VLV 7KH FRPSOHPHQWDU\ K\SRWKHVLV LV WKDW D GRPDLQ DQDO\VLV PHWKRG WKDW GRHV QRW ORRN IRU EHKDYLRUDO LQIOXHQFHV DQG WKH OLNH D VRIWH[FOXVLYH GRPDLQ DQDO\VLV

(80) LV D JRRG HQRXJK

(81) ILW IRU D VLP SOH GRPDLQVLPSOH SUHGLFWLRQ SUREOHP &KDSWHU  LQWURGXFHV UHTXLUHPHQWV IRU D SUHGLFWLRQ LQVWUXPHQW GHYHORS PHQW PHWKRG WKDW XVHV VRIWLQFOXVLYH GRPDLQ DQDO\VLV 7KH UHTXLUHPHQWV DUH WKHQ XVHG WR FUHDWH RXU PHWKRG 3UHGLFWLRQ ,QVWUXPHQW 'HYHORSPHQW IRU &RP SOH[ 'RPDLQV 7KH FKDSWHU DOVR GHVFULEHV D SUHGLFWLRQ LQVWUXPHQW GHYHORS PHQW PHWKRG ZLWK VRIWH[FOXVLYH GRPDLQ DQDO\VLV EDVHG RQ 6KPXHOL DQG .RS SLXV PHWKRG IRU SUHGLFWLYH DQDO\WLFV DQG :HEVWHU DQG :DWVRQ¶V VWUXFWXUHG OLW HUDWXUH UHYLHZ PHWKRG     %RWK PHWKRGV DUH YDOLGDWHG LQ FDVH VWXGLHV GHVFULEHG LQ &KDSWHUV   DQG  7KHVH FDVH VWXGLHV VHUYH QRW RQO\ DV YDOLGDWLRQ RI WKH PHWKRG LH LWV DSSOL FDELOLW\ LQ SUDFWLFH EXW DOVR DV YDOLGDWLRQ IRU WKH K\SRWKHVHV DERYH &KDSWHU  GHPRQVWUDWHV WKH XVH RI WKH VRIWH[FOXVLYH PHWKRG IRU D VLPSOH GRPDLQ WR VKRZ WKDW WKHUH LV LQGHHG D JRRG ILW LH VXIILFLHQW SUHGLFWLYH SRZHU 7R HQVXUH WKH VLPSOLFLW\ RI WKH GRPDLQ DQG WKDW LW LV QRW LQIOXHQFHG E\ DQ\ µVRIW¶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

(82) ,QWURGXFWLRQ. 6LPSOH 6RIWH[FOXVLYH. &KDSWHU . 'RPDLQ W\SH &RPSOH[ &KDSWHU  . $QDO\VLV W\SH 6RIWLQFOXVLYH. &KDSWHU . 7DEOH  $Q RYHUYLHZ RI WKH FDVH VWXGLHV LQ WKLV WKHVLV 7KH FDVH VWXGLHV FRYHU WKH PHDQ LQJIXO FRPELQDWLRQV RI WKH WZR W\SHV RI GRPDLQ VLPSOH DQG FRPSOH[

(83) DQG WKH WZR W\SHV RI GRPDLQ DQDO\VLV VRIWLQFOXVLYH DQG VRIWH[FOXVLYH

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

(85) i 6WUXFWXUH. .

(86)  0HWKRGRORJ\. 7KLV FKDSWHU EXLOGV RQ WKH WKHRU\ SUHVHQWHG LQ WKH SUHYLRXV FKDSWHU E\ JLYLQJ PHWKRGV IRU SUHGLFWLRQ LQVWUXPHQW GHYHORSPHQW 7KH ILUVW DQG PDLQ PHWKRG SUHVHQWHG KHUH XVHV D VRIWLQFOXVLYH GRPDLQ DQDO\VLV ZKLFK WDNHV EHKDYLRU SROLWLFV HWKLFV FXOWXUH DQG FRQIOLFWLQJ

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

(89) i 5HTXLUHPHQWV. i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

(90) 0HWKRGRORJ\. 4XDQWLWDWLYH DQG TXDOLWDWLYH PHWKRGV 7KH GRPDLQ DQDO\VLV SDUW RI WKH PHWKRG VKRXOG FDSWXUH VRIW IDFWRUV ZKLFK ILWV ZLWK WKH XVH RI TXDOLWDWLYH PHWKRGV VXFK DV LQWHUYLHZLQJ ,Q FRQWUDVW WKH GHYHORSPHQW RI SUHGLFWLYH PRGHOV XVHV TXDQ WLWDWLYH PHWKRGV VXFK DV VWDWLVWLFDO DQDO\VLV DQG PDFKLQH OHDUQLQJ 7KHUHIRUH WKH FRPELQDWLRQ RI WKHVH WZR SDUWV WKH SUHGLFWLRQ LQVWUXPHQW GHYHORSPHQW PHWKRG LV DQ H[DPSOH RI PL[HG PHWKRGV UHVHDUFK   ,Q SUHGLFWLRQ LQVWUX PHQW GHYHORSPHQW WKH TXDQWLWDWLYH PHWKRGV DUH XVHG WR YDOLGDWH WKH TXDOL WDWLYH PHWKRGV SUHGLFWLYH PRGHOV DUH FUHDWHG WR WHVW ZKLFK IDFWRUV IURP WKH GRPDLQ DQDO\VLV WUXO\ DIIHFW WKH SUHGLFWLRQ RXWFRPH ,W FRXOG DOVR EH VDLG WKDW WKH TXDQWLWDWLYH SUHGLFWLYH PRGHOV H[WHQG WKH NQRZOHGJH RI WKH GRPDLQ E\ VKRZLQJ ZKLFK IDFWRUV DIIHFW ZKDW LV SUHGLFWHG 7KLV W\SH RI UHODWLRQ EHWZHHQ TXDOLWDWLYH DQG TXDQWLWDWLYH PHWKRGV RU PL[HG PHWKRG SXUSRVH

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

(92)  ILQGLQJ RXW ZKHWKHU WKH GDWD LV PDQLSXODWHG RU QRLV\ IURP WKDW VDPH GDWD LV GLIILFXOW LI QRW LPSRVVLEOH  6HH 6HFWLRQ  IRU D PRUH GHWDLOHG GHVFULSWLRQ RI FRQVWUDLQWV DFWLRQDELOLW\ DQG WKHLU UHOD WLRQ. .

(93) i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ĆFLHQF\ $V D ILQDO UHTXLUHPHQW WKH SUHGLFWLRQ LQVWUXPHQW GHYHORSPHQW PHWKRG VKRXOG EH HIILFLHQW $V LW UHSODFHV LPSOLFLW GRPDLQ DQDO\VLV RU RWKHU RSDTXH PHWKRGV RI JDWKHULQJ GRPDLQ LQWHOOLJHQFH LW VKRXOG SURYLGH DV OLWWOH EXUGHQ DV SRVVLEOH 7KLV PHDQV WKDW QR XQQHFHVVDU\ SUHGLFWLYH PRGHOV VKRXOG EH GHYHORSHG. i 3UHGLFWLRQ LQVWUXPHQW GHYHORSPHQW IRU FRPSOH[ GRPDLQV ,Q WKLV VHFWLRQ ZH GHVFULEH RXU PHWKRG IRU SUHGLFWLRQ LQVWUXPHQW GHYHORS PHQW WKDW FDSWXUHV DQG XVHV VRIW IDFWRUV IURP WKH GRPDLQ FDOOHG SUHGLFWLRQ LQVWUXPHQW GHYHORSPHQW IRU FRPSOH[ GRPDLQV 3,'&'

(94)  2XU PHWKRG LV EDVHG RQ LQWHOOLJHQFH PHWDV\QWKHVLV ZKLFK KDV SUHYLRXVO\ EHHQ XVHG WR FUHDWH HFR QRPLF PRGHOV  DQG DV DQ DQDO\VLV WRRO IRU DVVRFLDWLRQ UXOH PLQLQJ   7KH QDPH GHULYHV IURP WKH IDFW WKDW LW V\QWKHVL]HV QHZ LQWHOOLJHQFH LQ WKH IRUP RI D PRGHO

(95) IURP GRPDLQ LQWHOOLJHQFH $V WKLV PHWKRG JDWKHUV DQG XVHV LQWHOOL JHQFH WR FUHDWH LQWHOOLJHQFH WKH DXWKRUV FRQVLGHU LW µPHWD¶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

(96) 0HWKRGRORJ\. JHQFH 7KH XQLTXH DQG LQWHUHVWLQJ K\SRWKHVHV DUH WKHQ GHWHUPLQHG LQ WKH K\ SRWKHVLV FRQYHUJHQFH VWHS ,Q VWDJH , WKH GRPDLQ H[SHUWV DUH DOVR FRQVXOWHG WR GHWHUPLQH WKH ILYH GRPDLQ FRQVWUDLQWV 7KH ZKROH RI VWDJH ,, LV EDVHG RQ 6KPXHOL DQG .RSSLXV¶ PHWKRG IRU SUHGLF WLYH DQDO\WLFV    DQG FRQVLVWV RI YDULDEOH VHOHFWLRQ GDWD FOHDQLQJ WHFK QLTXH VHOHFWLRQ DQG HYDOXDWLRQ ,Q VWDJH ,, SUHGLFWLYH PRGHOV DUH GHYHORSHG EDVHG RQ WKH GRPDLQ DQDO\VLV IURP VWDJH , 7KH K\SRWKHVHV DUH XVHG DV D EDVLV IRU YDULDEOH VHOHFWLRQ DQG GDWD FOHDQLQJ 7KH GDWD DQG GRPDLQ FRQVWUDLQWV DUH XVHG WR GHWHUPLQH ZKLFK K\SRWKHVHV FDQ EH WHVWHG D K\SRWKHVLV FDQ RQO\ EH WHVWHG LI WKHUH LV HQRXJK GDWD RI VXIIL FLHQW TXDOLW\ DYDLODEOH

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i 3UHSDUDWLRQ VWDJH. 3UREOHP LGHQWLĆFDWLRQ 3UHSDUDWLRQ IRU 3,'&' VWDUWV ZLWK DQ DVVHVVPHQW RI WKH SUREOHP GRPDLQ DQG GHILQLQJ WKH SUHGLFWLRQ JRDO µZKDW QHHGV SUHGLFWLQJ¶  S 

Referenties

GERELATEERDE DOCUMENTEN

Figuur 11 Voorkeursvariant met ontsluiting van de Zeeheldenwijk via de Karel Doormanweg Met de maatregelen van de voorkeursvariant ontstaat een netwerk van wegen waarbij

kandidaat iets verzwijgen, dan is dat achteraf eenduidig vast te stellen. In die zin gaat van de procedure een preventieve werking uit. Nieuw is dat de burgemeester en

vormgeving en een training voor rationeel rijden, kan de Renault Trucks D Optifuel uw brandstofverbruik aanzienlijk terugdringen en uw..

kraamverzor ger, medewerker thuiszorg waardoor zij in verband met de uitoefening van hun beroep of bedrijf meer dan tien keer per twee maanden aangewezen zijn op het gebruik van

Hvis rengøringsmaskinen efterlades hvor den kan blive udsat for frost, når arbejdet er overstået eller til opbevaring, skal der anvendes antifrostvæske for at forebygge

Op de ontwerp-instemmingsbesluiten zijn in totaal 31 zienswijzen binnengekomen (waarvan 29 uniek) en 7 reacties

De reden waarom de verzoeker fysiek blijvend niet in staat is deze negatieve wilsverklaring te schrijven en te ondertekenen is de volgende: .... werd aangewezen om deze

Om nu te be- grijpen op welke wijze het schrijven van de spiri- tuele autobiografie de deelnemers in staat stelt tot spirituele heroriëntatie en zelfbevestiging en