Prediction Off Pneumonia
Prediction Off Pneumonia
2021 11th International Conference on Cloud Computing, Data Science & Engineering (Confluence) | 978-1-6654-1451-7/20/$31.00 ©2021 IEEE | DOI: 10.1109/Confluence51648.2021.9377169
%DVHGRQ6WDWLVWLFDO$QDO\VLVRI0DWHUQDO&RQGLWLRQ
8VLQJ0DFKLQH/HDUQLQJ$SSURDFK
0G0HKHGL+DVDQ0G2PDU)DUXN%LGHVK%LVZDV%LNL
0G5LDMXOLLVODP.KDLUXO$ODP6\HGD)DUMDQD6KHWX
'HSDUWPHQWRI&RPSXWHU6FLHQFHDQG(QJLQHHULQJ
'DIIRGLO,QWHUQDWLRQDO8QLYHUVLW\'KDND%DQJODGHVK
כPHKHGL#GLXHGXEGRPDU#GLXHGXEGELGHVK#GLXHGXEG
ULDMXO#JPDLOFRPNKDLUXO#GLXHGXEGVKHWX#GLXHGXEG
$EVWUDFW² 3QHXPRQLD LV RQH RI WKH FRPPRQ GLVHDVHV ,FHFUHDP 3UHWHUP %LUWK ([FHVVLYH 6ZHDWLQJ &ROG RU
DPRQJVW FKLOGUHQ LQ %DQJODGHVK 0DQ\ FKLOGUHQ GLH IURP &RXJK:HKDYHIRXQGLQRXUVXUYH\WKDW%DWKLQJ7LPHDQG
SQHXPRQLD LQ %DQJODGHVK 3QHXPRQLD LV DQ LQIHFWLRQ WKDW &ROGRU&RXJKFDQDOVROHDGWKHEDELHVWRZDUGVSQHXPRQLD
LQIHFWVWKHDLUVDFVLQRQHRUERWKOXQJV,Q%DQJODGHVKQHDUO\ $IWHU H[DPLQLQJ RXU VXUYH\ ZH KDYH DSSOLHG PDFKLQH
FKLOGUHQ GLH RI SQHXPRQLD HYHU\ \HDU )RU GLVHDVHV
OHDUQLQJ DOJRULWKPV WR RXU FROOHFWHG GDWD :H NQRZ DW
IRUHFDVWLQJ0DFKLQHOHDUQLQJDOJRULWKPVDUHSRSXODUDQGXVHG
H[WHQVLYHO\ 0DFKLQH /HDUQLQJ DOORZV XV WR IXOILOO VXFK D WDVN SUHVHQW WLPH WKLV W\SH RI SUREOHP FDQ EH VROYHG XVLQJ WKH
ZLWK PXFK FRQVHTXHQFH :H HVWDEOLVKHG RXU GDWDVHW IURP WKH PDFKLQH OHDUQLQJ DSSURDFK .XUW %HQLUVFKNH HW DO >@
SDUWLFXODU REWDLQDEOH IURP RXU VXUYH\ )RU SURJQRVWLFDWLQJ VKRZHGWKDWFRSLRXVGLVHDVHVPD\KLQGHUSUHJQDQF\,QWKHLU
SQHXPRQLD ZH HPSOR\HG VL[ WUDGLWLRQDO 0DFKLQH /HDUQLQJ H[SHULHQFHZLWKVRPHRIWKHVHFRPSOLFDWLRQVRISUHJQDQF\
DOJRULWKPV7KH\DUH.1HDUHVW1HLJKERU .11 1DLYH%D\HV WKHSODFHQWDOMXGJPHQWVPD\XWWHUO\EHFRUURERUDWLYHRIWKH
FODVVLILHU 'HFLVLRQ 7UHH 6XSSRUW 9HFWRU 0DFKLQH 690 GLVHDVH0LFKDHO6.UDPHUHWDO>@LQWKHLUSDSHUVKRZHG
1HXUDO 1HWZRUN DOJRULWKP DQG 5DQGRP )RUHVW )RU WKDW WR HYDOXDWH WKH LPSDFWV RI SUHVFULELQJ DQ DQWLJHQ
LPSOHPHQWLQJWKHVHDOJRULWKPVZHDSSOLHG6FLNLWOHDUQ3DQGDV DYRLGDQFH GLHW ZKLOH SUHJQDQF\ RU ODFWDWLRQ RU ERWK RQ
1XP3\DQGIRUYLVXDOL]LQJRXUGDWDZHKDYHXVHG0DWSORWOLE
PRWKHUO\ DQG QHZERUQ QXWULWLRQ DQG RQ WKH SUHYHQWLRQ RU
DQGVHDERUQ%\SURSHULQWHUSUHWDWLRQZHFRQVLGHUHGWKHEHVW
SHUIRUPLQJDOJRULWKPIRUWKHSUHGLFWLRQRISQHXPRQLD:HKDYH UHPHG\ RI DWRSLF GLVHDVH LQ WKH FKLOG 0DWHUQDO GLHWDU\
PHDVXUHG WR FODVVLI\ ZKHWKHU SQHXPRQLD GHFOLQHV XQGHU DQWLJHQVDUHDOVRDFNQRZOHGJHGWRVSDQWKHSODFHQWD,QWKLV
SQHXPRQLD 3RVLWLYH DQG SQHXPRQLD 1HJDWLYH FODVV $PRQJ DJH RI DUWLILFLDO LQWHOOLJHQFH PDFKLQHV DFW OLNH D KXPDQ
DOOWKHDOJRULWKPVZHKDYHFKRVHQWKHEHVWDOJRULWKPZKLFKLV )RUPLQJ D SUHGLFWLRQ RI SQHXPRQLD DQG VWDWLVWLFDO
SURYLGHG XV EHVW DFFXUDF\ DQG )VFRUH %\ WKH EHVW H[DPLQDWLRQRIPDWHUQDOILWQHVVLVQRWDVLPSOHWDVN0DFKLQH
DFFRPSOLVKLQJ DOJRULWKP RXU PRGHO FDQ SUHGLFW SQHXPRQLD OHDUQLQJLPSURYHVXVE\JLYLQJRSWLPDOIRUHFDVWH[SHUWLVHOLNH
TXLWHZHOO UHLQIRUFHPHQW OHDUQLQJ DOJRULWKPV VXSHUYLVHG OHDUQLQJ
DOJRULWKPVDQGXQVXSHUYLVHGOHDUQLQJDOJRULWKPV$VZHDUH
.H\ZRUGV²0DFKLQH /HDUQLQJ 'HFLVLRQ WUHH 3QHXPRQLD
6WDWLVWLFDODQDO\VLV'LVHDVHGHWHFWLRQ
ZRUNLQJ RQ SUHGLFWLRQ ZH FKRRVH VXSHUYLVHG OHDUQLQJ
DOJRULWKPV WR PDNH IRUHFDVWLQJ 7KH JLYHQ 0/ WHFKQLTXHV
, ,1752'8&7,21 FDQEHXVHGIRUSUHGLFWLQJSQHXPRQLDDQGVWDWLVWLFDODQDO\VLV
3QHXPRQLD LV WKH PRVW PRPHQWRXV RXWFRPH RI DFXWH RIWKHPDWHUQDOFRQGLWLRQ7KLVLVWKHDLPRIRXUZRUN
UHVSLUDWRU\LQIHFWLRQ $5, DQGNLOOVPRUHFKLOGUHQWKDQDQ\
RWKHUFRQWDJLRXVGLVHDVH,WLVRQHRIWKHYHU\XVXDOGLVHDVHV 7KLV SDSHU ZLOO EH DUUDQJHG DV IROORZV ,Q 6XEGLYLVLRQ ,,
DPRQJQHZERUQEDELHV(YHQGXULQJSUHJQDQF\EDELHVFDQ UHYLHZRIUHODWHGZRUNLVJLYHQ,QVHJPHQW,,,WKHSURSRVHG
EH DIIHFWHG E\ SQHXPRQLD $OPRVW HYHU\ \HDU PHWKRGRORJ\LVSURYLGHG,QXQLW,9ZHILQGWKHH[SHULPHQWDO
FKLOGUHQ XQGHU ILYH \HDUV RI DJH RU DURXQG HYHU\GD\ UHVXOWV DQG FRPSDUH GLIIHUHQW UHVXOWV WKDW ZH DFKLHYHG
ORVVHVWKHLUOLYHV$PRQJWKHPRYHUDUHQHZERUQV WKURXJKGLIIHUHQW0/WHFKQLTXHV,QVHFWLRQ9ZHFRQFOXGHG
ZKLFK DUH DURXQG RI WKH WRWDO GHDWK UDWH >@ ,Q WKH RXUPRGH)LQDOO\ZHFRQFOXGHGRXUZRUNZLWKDFRQFOXVLRQ
FRQWH[WRI%DQJODGHVKSQHXPRQLDLVUHVSRQVLEOHIRUDOPRVW FRQWDLQLQJIXWXUHZRUNVLQVHFWLRQ9,
RIWKHGHDWKVRIFKLOGUHQXQGHUILYH\HDUVRIDJHZKLFK
LVDSSUR[LPDWHO\>@6WDWLVWLFVKDYHVKRZQWKDWPRVW ,, /,7(5$785(5(9,(:
RIWKHSQHXPRQLDFDVHVKDYHRFFXUUHGLQYLOODJHV:HKDYH
GRQH RXU VXUYH\ RQ ZRPHQ LQ WKH YLOODJH ,Q RXU
TXHVWLRQQDLUHZHKDYHREVHUYHGWKDWDJHLVRQHRIWKHPDLQ &XUUHQWO\WKHXVHRI0DFKLQHOHDUQLQJLVYDVWLQDOPRVWHYHU\
IDFWRUV FDXVLQJ SQHXPRQLD <RXQJHU ZRPHQ GXULQJ FDVH,WKDVDOVREURXJKWDORWVRIVXFFHVVLQPHGLFDOVFLHQFH
SUHJQDQF\DUHPRUHOLNHO\WRKDYHSQHXPRQLDLQWKHLUEDELHV 3UHGLFWLRQ RI GLIIHUHQW GLVHDVH LV RQH RI WKH PRVW SRSXODU
$OOHUJLF5KLQLWLVLVDOVRUHVSRQVLEOHIRUFDXVLQJSQHXPRQLD VHFWRU,WPDNHVRXUOLIHHDV\DQGUHDVRQDEOH6RPHRIUHODWHG
7KHUHDUHDOVRRWKHUFLUFXPVWDQFHVOLNH%DWKLQJ7LPH(DWLQJ ZRUNLVJLYHQEHOORZ
978-1-6654-1451-7/21/$31.00 2021
c IEEE 919
Authorized licensed use limited to: Rutgers University. Downloaded on May 17,2021 at 13:58:31 UTC from IEEE Xplore. Restrictions apply.
)HQJHWDO>@ILQGRXWDUHWURVSHFWLYHPXOWLFHQWHUVWXG\RI ILQG WKH EHVW DOJRULWKP )LUVW WKH\ XVHG DOO DWWULEXWHV WR
SDWLHQWV ZKR KDYH PLOG &RYLG SQHXPRQLD WR DVVHVV WKH SUHGLFW WKHQ UHPRYH VRPH DWWULEXWHV WR ILQG RXW WKH EHVW
HIILFDF\ RI FKHVWFDOFXODWHG WRPRJUDSK\ ZLWK FOLQLFDO DOJRULWKP'HFLVLRQWUHH1DLYH%D\HV/RJLVWLF5HJUHVVLRQ
FKDUDFWHULVWLFV IRU ULVNVWUDWLI\ SDWLHQWV 7KH\ KDYH XVHG 5DQGRP )RUHVW 690 .11 DOJRULWKP XVHG IRU SUHGLFWLRQ
ORJLVWLF UHJUHVVLRQ DQG QRPRJUDP HVWDEOLVKPHQW WR DQDO\]H DQG'HFLVLRQWUHHJLYHVWKHEHVWUHVXOWZLWKWKHDFFXUDF\RI
SDWLHQW¶VGDWD
<DQTLX HW DO >@ GHVFULEHG D ZRUN WKDW DLPV DW H[DPLQLQJ
PRGHOV IRU SRVWVWURNH SQHXPRQLD SUHGLFWLRQ SDUWLFXODUO\ ,,, 0(7+2'2/2*<
GHHS OHDUQLQJ DSSURDFKHV XVLQJ PRUH DGYDQFHG PDFKLQH )RU FRPSOHWLQJ RXU ZRUN ZH IROORZHG WKHVH VWHS
OHDUQHGDOJRULWKPV7KH\XVHGWUDLQVHWDQGVHWWR FRUUHVSRQGLQJO\
HYDOXDWH WKHLU ZRUN 7KH\ XVHG GLIIHUHQW PDFKLQH OHDUQLQJ
DQG 1HXUDO QHWZRUN 0/3 DOJRULWKPV 7KH\ VHOHFWHG
5(7$,1PHWKRGIRUSUHGLFWLRQ%HFDXVHLWSURGXFHGEHWWHU
SHUIRUPDQFH ZLWK $8& VHQVLWLYLW\ 6SHFLILFLW\
3DXOD'LHKUHWDO>@GHVFULEHWKDWFRXJKZDVFRQVLGHUHGWKH 'DWD&ROOHFWLRQ
ILIWKPRVWFRPPRQFDXVHIRUGRFWRU VYLVLWVEXWNQRZOHGJH
DERXWVHULRXVFRXJKZDVUDUHO\UHTXHVWHGRQDWUHQGDQGZDV
QRW GLVVHFWHG GXULQJ PXOWLYDULDWH WUHDWPHQW 7KH\ DOVR
UHJLVWHUHG&RXJKSDWLHQWVDOOZLWKDJUDGHGKLVWRU\D
3UHSURFHVVLQJ
SK\VLFDOKLVWRU\DQGDQ;UD\OXQJ-XVW SHUFHQW RI
WKHVH ZHUH IRXQG WR KDYH FULWLFDO ;UD\ LQILOWUDWLRQ
SQHXPRQLD 7KH\ KDYH D JXLGHOLQH WKDW VHSDUDWHV
SQHXPRQLDSDWLHQWVZLWKDVHQVLWLYLW\RIDQGRUD
VHQVLELOLW\RIDQGDSUHFLVLRQRI $OJRULWKPV
6WDWLVWLFDO$QDO\VLV
0RKDPHGHWDO>@DOVRWULHGWRLQYHVWLJDWHWKHVHSDUDWLRQRI ,PSOHPHQWDWLRQ
EDFWHULDODQGYLUDOFRQGLWLRQVLQSHGLDWULFSRSXODWLRQDFTXLUHG
SQHXPRQLD E\ & 5HDFWLYH 3URWHLQ &53 QHXWURSKLO DQG
ZKLWH FHOO FRXQWV DV D EDVLV WR FUHDWH FXWV RI SURYRFDWLYH
ODEHOLQJDQG :&& ,QLQIDQWVEDFWHULDOSQHXPRQLDLV (YDOXDWLRQRI0RGHO
ROGHU WKDQ YLUDO SQHXPRQLD FKLOGUHQ %DFWHULDO GLVHDVHV
DVVRFLDWHGWRYLUXVHVKDYHDKLJKHUIUHTXHQF\RIPHGLDQ&53
WKDQ :&& DQG QHXWURSKLO FRXQWV 7KH 52& FXUYH RI WKH
SURWRW\SHIRUGLVFULPLQDWLQJEDFWHULDOIURPYLUDOSQHXPRQLD
EDVHG RQ DJH &53 DQG QHXWURSKLO FDOFXODWLRQ SURYLGHG )LJ0HWKRGRORJ\GLDJUDP
H[SDQVHXQGHUWKHFXUYHRIZLWKVHQVLWLYLW\DQG $ 'DWD&ROOHFWLRQ
VSHFLILFLW\
1LQJHWDO>@UHSUHVHQWHGDQRSHQUHVRXUFHFRPSULVLQJGDWD :H FROOHFWHG RXU GDWD E\ VXUYH\LQJ WKH UXUDO DUHD RI
IURP SDWLHQWV ZLWK SQHXPRQLD LQFOXGLQJ &29,' %DQJODGHVK 7KLV VXUYH\ ZDV SHUIRUPHG RQO\ IRU :RPHQ
SQHXPRQLD FRQVLVWLQJRIFKHVWFRPSXWHGWRPRJUDSK\ &7 ZKRJDYHELUWKEHWZHHQWR:HVHOHFWHG1DPH$JH
LPDJHV FOLQLFDO IHDWXUHV IURP D UDQJH RI ELRFKHPLFDO /RFDWLRQ LV FRPPRQ IRU DOO DQG WKH SDUDPHWHUV WKDW DUH
DQG FHOOXODU DQDO\VHV RI EORRG DQG XULQH VDPSOHV DQG UHVSRQVLEOH IRU RFFXUULQJ 3QHXPRQLD RI D QHZERUQ EDE\
ODERUDWRU\FRQILUPHG VHYHUH DFXWH UHVSLUDWRU\ V\QGURPH 7KHVH W\SHV RI SDUDPHWHUV DUH $OOHUJLF 5KLQLWLV %DWKLQJ
FRURQDYLUXV 6$56&R9 FOLQLFDO VWDWXV 7KH\ DOVR 7LPH (DWLQJ ,FH &UHDP 3UHWHUP ELUWK DQG ([FHVVLYH
VKRZHG WKH HIILFLHQF\ RI WKH GDWDEDVH IRU WKH SURSKHF\ RI 6ZHDWLQJ2XUZRUNLVQRWRQO\DSQHXPRQLDSUHGLFWLRQEXW
&29,' PRUELGLW\DQG PRUWDOLW\RXWFRPHVXVLQJDGHHS DOVR D 6WDWLVWLFDO DQDO\VLV RI WKHVH SDUDPHWHUV WKDW FDXVHV
OHDUQLQJDOJRULWKPWUDLQHGZLWKGDWDIURPSDWLHQWVDQG 3QHXPRQLD 6R HDFK RI RXU VHOHFWHG DWWULEXWHV LV YHU\
PDQXDOO\ ODEHOHG &7 VOLFHV ,Q D QRQSDUWLVDQ LPSRUWDQW IRU RXU ZRUN 6R ZH FDUHIXOO\ FROOHFWHG WKHVH
YDOLGDWLRQJURXSRISDWLHQWVWKHDOJRULWKPGLVWLQJXLVKHG SDUDPHWHUVE\DTXHVWLRQQDLUHWKDWLVEHOORZHG
DPLGVW QHJDWLYH PLOG DQG FULWLFDO SUHGLFDPHQWV ZLWK ILHOGV
EHQHDWKWKHUHFHLYHUFRQWDLQLQJRULJLQDOFXUYHRI :DVVKHDIIHFWHGE\$OOHUJLF5KLQLWLVGXULQJSUHJQDQF\"
DQGVHTXHQWLDOO\ :KDWZDV\RXU%DWKLQJ7LPH" 0RUQLQJ$IWHUQRRQ
$VZDWKL $ . HW DO >@ KDYH SURSRVHG D PHWKRG IRU :DVVKHHDWHQ,FH&UHDPUHJXODUO\GXULQJSUHJQDQF\"
FODVVLILFDWLRQ DQG GLDJQRVLV RI WK\URLG GLVHDVH WR GHWHFW LWV :KDWZDVKLVGHOLYHU\WLPH" 3UHWHUPRUQDWXUDOWLPH "
HDUO\ VWDJH XVLQJ :HLJKWHG 690 FODVVLILFDWLRQ DQG WR :DVVKHDIIHFWHG([FHVVLYH6ZHDWLQJSUREOHP"
RSWLPL]H 690 SDUDPHWHUV VXFK DV 76+ 7 7 WKH\ XVH :DVKHUEDE\DIIHFWHGE\3QHXPRQLD"
SDUWLFOH VZDUP RSWLPL]DWLRQ $OVR WKH\ XVH .11 WR
DSSUR[LPDWHWKHPLVVLQJYDOXHYLDXVHULQSXW7KLVSURSRVHG 0RVWRIWKHTXHVWLRQVZHUH\HVQRW\SH
ZRUN FRQVWUXFWHG WKH 3UHGLFWLRQ RI 7K\URLG 'LVHDVH 8VLQJ % 3UHSURFHVVLQJ
'DWD0LQLQJ7HFKQLTXHV
'DWD FOHDQLQJ 2XU LQLWLDOO\ FROOHFWHG GDWD FRQWDLQV
5-DQH 3UHHWKD 3ULQF\ HW DO >@ KDYH SUHGLFWHG FDUGLDF
1DPH DQG /RFDWLRQ WKHVH DUH XQQHFHVVDU\ IRU WUDLQLQJ RXU
GLVHDVH XVLQJ VXSHUYLVHG PDFKLQH OHDUQLQJ DOJRULWKPV DQG
920 2021 11th International Conference on Cloud Computing, Data Science & Engineering (Confluence 2021)
Authorized licensed use limited to: Rutgers University. Downloaded on May 17,2021 at 13:58:31 UTC from IEEE Xplore. Restrictions apply.
PRGHO6RZHHOHPHQWDOOQDPHVDQGORFDWLRQVIURPWKHPDLQ RI WKH WLPH GHOLYHU\ UDWH GHFUHDVLQJ ZKHQ WKH DJH LV
GDWDVHW LQFUHDVLQJ
7UDQVIRUPDWLRQ 7R DFTXLUH TXDOLW\ NQRZOHGJH GDWD
WUDQVIRUPDWLRQLVDYHU\LPSRUWDQWSDUWRIDQ\GDWDVHW>@
PDFKLQHOHDUQLQJDOJRULWKPVGRQRWXQGHUVWDQGVWULQJVRZH
QHHGWRFRQYHUWDOOVWULQJWRDQXPEHU)RUWKLVSXUSRVHIRU
\HVZHXVHGDQGIRU1R
)HDWXUH 6HOHFWLRQ )RU WKH LQGHSHQGHQW YDULDEOH ZH
VHOHFWHG $JH $OOHUJLF 5KLQLWLV %DWKLQJ 7LPH (DWLQJ ,FH
&UHDP 3UHWHUP ELUWK DQG ([FHVVLYH 6ZHDWLQJ $QG &KLOG
3QHXPRQLD LV D GHSHQGHQW YDULDEOH WKDW LV SUHGLFWHGE\ RXU
PRGHO
)LJ$OOFROOHFWHGDJHV
,WLVDJUHDWSUREOHPLQRXUFRXQWU\:HQHHGWRRYHUFRPHWKLV
SUREOHPE\LQFUHDVLQJDZDUHQHVVDQGWKHULVNWRJLYHHDUO\
ELUWK7KHOLJKWRIKRSHLVLQRXUFRXQWU\ORWVRIRUJDQL]DWLRQV
OLNH :5$ %DQJODGHVK ,QWHUQDWLRQDO &HQWUH IRU 'LDUUKRHDO
'LVHDVH5HVHDUFK%DQJODGHVK ,&''5% ZRUNWRVROYHWKLV
W\SH RI SUREOHP 5HFHQWO\ ,&''5% KDV GRQH ORWV RI
YDOXDEOHUHVHDUFKDERXWPDWHUQDODQGFKLOGKRRGPDOQXWULWLRQ
)LJ+HDWPDSWRFRUUHODWLRQ
)LJ 5HSUHVHQW WKH FRUUHODWLRQ ZLWK HDFK DQG HYHU\
DWWULEXWH ,Q WKLV JUDSK GHHSHU FRORU UHSUHVHQWV OHVV
FRUUHODWLRQDQGOLJKWHUFRORUUHSUHVHQWVWKHKLJKHVWFRUUHODWLRQ
EHWZHHQ WZR DWWULEXWHV 7KH KLJKHVW FRUUHODWLRQ ZLWK
SQHXPRQLDLVIRUWKHDWWULEXWH3UHWHUPELUWKORZHVWLV
IRUWKHDWWULEXWH$JHV
& 6WDWLVWLFDO$QDO\VLV
,QWKLVSRUWLRQZHWULHGWRILQGRXWWKHFDXVHVRISQHXPRQLD
LQDQHZERUQEDE\,VLWGHSHQGLQJRQWKHPRWKHU VKHDOWKRU
GDLO\ URXWLQH" )RU WKLV SXUSRVH ZH IRFXVHG RQ RXU
LQGHSHQGHQWDWWULEXWHV
$JH$QDO\VLV:HVXUYH\HGEHWZHHQWKHDJHVRIWR
DQGZHFROOHFWHGWKHDJHZKHQWKH\KDYHJLYHQELUWKWRD )LJ3QHXPRQLDUDWLRDFFRUGLQJWRPRWKHU¶VDJH
FKLOG)LJUHSUHVHQWVWKHDJHZKHQWKH\JDYHELUWKWRWKHLU
)URP RXU VXUYH\ ZH IRXQG SQHXPRQLD LQIDQWV RXW RI
EDE\RQWKH[D[LVDQGWKH \D[LVUHSUHVHQWVWKH QXPEHURI
)LJ,OOXVWUDWHVWKHHIIHFWRIPRWKHUDJHIRUFDXVH¶V
WKLVLQGHFHQW)URPWKLVJUDSKZHFDQVHHWKHKLJKHVWQXPEHU
SQHXPRQLD :H FDQ VHH WKH KLJKHVW SHUFHQWDJH RI DIIHFWHG
ZKHQ WKH DJH RI WKH\ KDYH JLYHQ ELUWK 7KH ORZHVW
UDWH RI SQHXPRQLD IRU QHZERUQ EDELHV DOO PRWKHUV
QXPEHUDURXQGZKHQWKHLUDJHZDV\HDUVROG6R]RXHW
ZKRKDYHWDNHQEDE\DWWKHDJHRI\HDUV$QGWKHORZHVW
DO>@KDYHIRXQGWKDWZRPHQKDYHKLJKIHUWLOLW\ZKHQWKH DIIHFWHGUDWHLVRIQHZERUQEDELHVZKHQWKHLUPRWKHU V
DJHRIDQGSUHJQDQF\ULVNVROHVV$QGIHUWLOLW\EHFRPHV DJHZDV\HDUV)RUPWKLVJUDSKLIZHFRPSDUHWKHDJHRI
GHFOLQHVDWWKHDJHRI$FFRUGLQJWR:+2>@WDNHD DQG\HDUVZHFDQVHHWKDWIURPWR\HDUV
EDE\ XQGHU \HDUV ROG LV DKLJKULVN ORZ ELUWK UDWH DQG D ROGPRWKHU¶VEDELHVOHVVDIIHFWHGWKDQWKHDJHRI\HDUV
KLJK FKDQFH RI SUHWHUP GHOLYHU\ ,I ZH KLJKO\ FULWLFL]H WKH ,PSDFW RI 5RXWLQHDQG 'LVHDVH 3UHJQDQF\ WLPH LV D
JUDSKZHFDQVD\WKDWLQRXUFRXQWU\7KHWHQGHQF\WRJLYH FUXFLDOPRPHQWIRUHYHU\ZRPDQ'XULQJWKLVWLPHLIVKHGRHV
ELUWKDWDQHDUO\DJHLVPXFKKLJKHULQ%DQJODGHVK$QGPRVW QRWIROORZDURXWLQHSURSHUO\WKHUHPLJKWEHVRPHEDGLPSDFW
2021 11th International Conference on Cloud Computing, Data Science & Engineering (Confluence 2021) 921
Authorized licensed use limited to: Rutgers University. Downloaded on May 17,2021 at 13:58:31 UTC from IEEE Xplore. Restrictions apply.
RQ KHU EDE\ 'LIIHUHQW GLVHDVH SUHJQDQF\ WLPH LV DQRWKHU LPSOHPHQWHG ZLWK WKH KHOS RI WKH PRVW SRSXODU 0/ OLEUDU\
IDFWRUWKDWFDXVHVDEDGHIIHFWRQWKHEDE\ VLFNLWOHDUQ7RDFTXLUHWKHEHVWDFFXUDF\ZHXVHGDGLIIHUHQW
SDUDPHWHUWKDWLVJLYHQLQ7DEOH,:HVHWWKHEHVWSDUDPHWHU
IRU HDFK DOJRULWKP IRU ZKLFK SDUDPHWHU WKH DOJRULWKP
SURGXFHVWKHKLJKHVWDFFXUDF\DQGEHVW)VFRUH
,9 (;3(5,0(17$/5(68/76
:H KDYH XVHG VHYHUDO 0DFKLQH /HDUQLQJ DOJRULWKPV WR
SUHGLFWSQHXPRQLD:HDSSOLHG.1HDUHVW1HLJKERU .11
1DLYH %D\HV FODVVLILHU 'HFLVLRQ 7UHH 6XSSRUW 9HFWRU
0DFKLQH 690 1HXUDO 1HWZRUN DOJRULWKP DQG 5DQGRP
)RUHVWIRUSUHGLFWLRQ:HFRPSDUHGWKHVHDOJRULWKPVEDVHG
RQDFFXUDF\DQG)VFRUH)URPWKHDFFXUDF\WDEOH7DEOH,,
:HFDQVHHWKDWIRUWHVWGDWDXVDJHUDWHZHJRW
IRU 1DLYH %D\HV IRU 'HFLVLRQ 7UHH IRU
690 IRU 1HXUDO 1HWZRUN IRU 5DQGRP
)RUHVWDQGWHVWGDWDXVDJHUDWHZHJRWDFFXUDF\
LQ.11
)URPWKH)VFRUHPDWUL[WDEOH7DEOH,,,:HDOVRVHHWKDWIRU
)LJ3RVLWLYHDQGQHJDWLYHUDWHRIDIIHFWHG WKH WHVW GDWD XVDJH UDWH ZH JRW IRU .11
IRUQDLYH%D\HVIRU'HFLVLRQ7UHH
7RNHHSLWLQPLQGZHDOVRDQDO\]HVRPHFRPPRQGLVHDVHV IRU 690 IRU 1HXUDO 1HWZRUN DQG IRU
DQGGDLO\DFWLYLWLHVRIZRPHQ)LJ5HSUHVHQWVSRVLWLYHDQG 5DQGRP)RUHVW&RQVLGHULQJUHVXOWVIURPDOOWKHDOJRULWKPV
QHJDWLYH DWWULEXWHV RI ZRPHQ ZKRVH EDE\ ZDV DIIHFWHG E\ ZHFKRVHWKH'HFLVLRQ7UHHIRULWVRXWVWDQGLQJSHUIRUPDQFH
SQHXPRQLD,QWKLV JUDSKEOXHFRORULQGLFDWHVSRVLWLYHDQG
RUDQJH FRORU LQGLFDWH QHJDWLYH $ERXW QXPEHUV RI WKH
PRWKHUDIIHFWHGE\FROGFRXJKZKRVHEDE\ZDVDIIHFWHGE\
SQHXPRQLD %DWKLQJ WLPH LV DOVR DQRWKHU UHDVRQ WKDW PD\
FDXVH SQHXPRQLD LQ LQIDQWV 2XU VXUYH\ UHSRUW VKRZV WKDW
PRVWRIWKHPRWKHUVDURXQGZKRZHUHEDWKLQJODWH DIWHU
30 WKHLUFKLOGUHQZHUHPRUHDIIHFWHGE\SQHXPRQLD
$OVRZHKDYHVHHQWKDWSUHWHUPELUWKWKHFDXVHRISQHXPRQLD
VDPH DV EDWKLQJ WLPH $URXQG ZRPHQ ZHUH DIIHFWHG
WKURXJK DOOHUJLF UKLQLWLV GXULQJ SUHJQDQF\ DQG WKHLU EDELHV
ZHUHDIIHFWHGE\SQHXPRQLDWKRXJKWKHQRWDIIHFWHGQXPEHUV
RI ZRPHQ ZHUH D OLWWOH ELW KLJKHU WKDQ WKUHH SUHYLRXV
DWWULEXWHV$WODVWZHFDQVHHWKHWHQGHQF\WRJHWSQHXPRQLD
RIQHZERUQEDELHVLVPXFKOHVVZKLOHWKHPRWKHUVHDWLQJLFH
FUHDPDQGH[FHVVLYHVZHDWLQJGXULQJSUHJQDQF\
' $OJRULWKPV,PSOHPHQWDLRQ
7$%/(, $775,%87(86$*(6
922 2021 11th International Conference on Cloud Computing, Data Science & Engineering (Confluence 2021)
Authorized licensed use limited to: Rutgers University. Downloaded on May 17,2021 at 13:58:31 UTC from IEEE Xplore. Restrictions apply.
7$%/(,, $&&85$&<7$%/(
$OJRULWKPV
7HVWLQJ
'DWDXVDJH
1HXUDO1HWZRUN 5DQGRP
UDWH .11 1DLYH%D\HV 'HFLVLRQWUHH 690 )RUHVW
7$%/(,,, )6&25(0$75,;7$%/(
$OJRULWKPV
7HVWLQJ
'DWDXVDJH
UDWH
1HXUDO1HWZRUN 5DQGRP
.11 1DLYH%D\HV 'HFLVLRQWUHH 690 )RUHVW
5HFDOOUDWHIRU1HJDWLYH
்ே ଶ
்ேାி ଶାଶ
)LJUHSUHVHQWVFRQIXVLRQPDWUL[RI'HFLVLRQWUHHDOJRULWKP
)URPDERYHFDOFXODWLRQZHFDQVHHWKDWRXUPRGHOLVEHWWHUIRU
SRVLWLYHSQHXPRQLDSUHGLFWLRQDVSRVLWLYHUHFDOOUDWHLVEHWWHU
WKDQQHJDWLYHUHFDOOUDWH
9, &21&/86,21$1')8785(:25.
(YHU\QHZERUQEDE\LVWKHIXWXUHRIWKLVZRUOG,WLVRXU
GXW\WRSURWHFWWKHLUKHDOWK,IZHFDQGHYHORSDV\VWHPWKDW
FDQ UHGXFH WKH LQFLGHQFH RI SQHXPRQLD LW ZLOO EH RXU JUHDW
VXFFHVV ,Q RXU VXJJHVWHG PRGHO ZH KDYH PDQLSXODWHG VL[
SRSXODU FODVVLILFDWLRQ 0/ DOJRULWKPV WR DFKLHYH WKH EHVW
RXWFRPH$IWHUHPSOR\LQJWKHVHDOJRULWKPVWRWKHGDWDVHWZH
)LJ&RQIXVLRQPDWUL[ SHUFHLYHGWKDWWKHH[DFWQHVVSURGXFHGE\WKHDOJRULWKPVZDV
QRW DOLNH 7KH GHFLVLRQ WUHH DOJRULWKP SHUIRUPHG WKH EHVW
DFFXUDF\ &RPSDUDWLYHO\ WKH 'HFLVLRQ WUHH JRW WKH EHVW
்ା்ே ଶାଶଷ
$FFXUDF\ DFFXUDF\ZHDSSOLHGWKLVIRUIRUHFDVWLQJRISQHXPRQLDGLVHDVH
்ାிାிேା்ே ଶଷାଶାାଶ
RIDQHZERUQEDE\DQGWKHVWDWLVWLFDODQDO\VLVRIWKHPDWHUQDO
FRQGLWLRQ,WLVGHVLUDEOHWRPLQLPL]HWKHGHDWKRILQIDQWV7KLV
ZRUNFDQDVVLVWWKHJRYHUQPHQWWRDZDUHWKHSHRSOH
(UURU
:HVSHFXODWHGWRDFFRPSOLVKWKHEHVWREWDLQDEOHFRQVHTXHQFH
EXW WKHUH DUH VWLOO D IHZ EDUULHUV WR RXU ZRUN 2QH RI WKH
5HFDOOUDWHIRUSRVLWLYH VLJQLILFDQWOLPLWDWLRQVRIRXUZRUNLVWKDWZHKDYHQ WVXUYH\HG
் ଶଷ
WKHZKROHFRXQWU\:HKDYHLQVSHFWHGVL[YLOODJHVRI&RPLOOD
்ାிே ଶଷା
6LUDMJDQMDQG*RSDOJDQM$GGLWLRQDOO\ZHKDYHZRUNHGZLWK
SUHJQDQW ZRPHQ RQO\ ,Q WKH IXWXUH :H ZLOO GHYHORS DQ
2021 11th International Conference on Cloud Computing, Data Science & Engineering (Confluence 2021) 923
Authorized licensed use limited to: Rutgers University. Downloaded on May 17,2021 at 13:58:31 UTC from IEEE Xplore. Restrictions apply.
LQWHOOLJHQWV\VWHPWKDWZLOOSUHGLFWSQHXPRQLDDQGUHGXFHWKH FRXJK²$ VWDWLVWLFDO DSSURDFK -RXUQDO RI &KURQLF
GHDWKUDWHRIQHZERUQEDELHVVPDUWO\%\XVLQJWKLVV\VWHP 'LVHDVHV´ 9ROXPH ,VVXH 3DJHV
GXULQJSUHJQDQF\SHRSOHZLOONQRZZKDWWKH\VKRXOGGRDQG ,661 KWWSVGRLRUJ
ZKDW VKRXOG QRW $QG DOVR ZH ZLOO WU\ WR HQKDQFH RXU GDWD
FROOHFWLRQE\JDWKHULQJGDWDIURPWKHZKROHFRXQWU\DQGZLOO
GHYHORSDQDQGURLGDSSOLFDWLRQZKLFKFDQEHHDV\WRXVHIRU > 0RKDPHG $(OHPUDLG 6WHSKHQ3 5XVKWRQ0DWWKHZ )
DQ\SHRSOH 7KRPDV'DYLG$6SHQFHU$QGUHZ5*HQQHU\-XOLD(
&ODUN 8WLOLW\ RI LQIODPPDWRU\ PDUNHUV LQ SUHGLFWLQJ WKH
DHWLRORJ\ RI SQHXPRQLD LQ FKLOGUHQ 'LDJQRVWLF
0LFURELRORJ\DQG,QIHFWLRXV'LVHDVH9ROXPH,VVXH
5()(5(1&(6 3DJHV ,661
KWWSVGRLRUJMGLDJPLFURELR
>@8QLFHI 2FWREHU 3QHXPRQLD $YDLODEOH >@1LQJ:/HL6<DQJ-HWDO2SHQUHVRXUFHRIFOLQLFDO
KWWSVGDWDXQLFHIRUJWRSLFFKLOG GDWD IURP SDWLHQWV ZLWK SQHXPRQLD IRU WKH SUHGLFWLRQ RI
KHDOWKSQHXPRQLDVWKDVKU/0:$NR%GSXI &29,'RXWFRPHVYLDGHHSOHDUQLQJ1DW%LRPHG(QJ
>@,&''5% 3QHXPRQLD DQG RWKHU UHVSLUDWRU\ GLVHDVHV ± KWWSVGRLRUJV
$YDLODEOH KWWSVZZZLFGGUERUJQHZVDQG
HYHQWVSUHVVFRUQHUPHGLDUHVRXUFHVSQHXPRQLDDQG >@$VZDWKL$.$QLO$QWRQ\³$Q,QWHOOLJHQW6\VWHPIRU
RWKHUUHVSLUDWRU\GLVHDVHV 7K\URLG 'LVHDVH &ODVVLILFDWLRQ DQG 'LDJQRVLV´ QG
>@.%HQLUVFKNH*-%XUWRQDQG51%DHUJHQ0DWHUQDO ,QWHUQDWLRQDO &RQIHUHQFH RQ ,QYHQWLYH &RPPXQLFDWLRQ
GLVHDVHV FRPSOLFDWLQJ SUHJQDQF\ GLDEHWHV WXPRUV DQG &RPSXWDWLRQDO 7HFKQRORJLHV ,&,&&7
SUHHFODPSVLD OXSXV DQWLFRDJXODQW LQ 3DWKRORJ\ RI WKH SS
KXPDQSODFHQWD6SULQJHUSS >@5-33ULQF\63DUWKDVDUDWK\36+HQF\-RVH$5DM
>@06.UDPHUDQG5-(%&+$&5-.DNXPD /DNVKPLQDUD\DQDQ DQG 6 -HJDQDWKDQ 3UHGLFWLRQ RI
0DWHUQDOGLHWDU\DQWLJHQDYRLGDQFHGXULQJSUHJQDQF\RU &DUGLDF 'LVHDVH XVLQJ 6XSHUYLVHG 0DFKLQH /HDUQLQJ
ODFWDWLRQRUERWKIRUSUHYHQWLQJRUWUHDWLQJDWRSLFGLVHDVH $OJRULWKPV WK ,QWHUQDWLRQDO &RQIHUHQFH RQ
LQWKHFKLOGYROQRSS ,QWHOOLJHQW &RPSXWLQJ DQG &RQWURO 6\VWHPV ,&,&&6
0DGXUDL ,QGLD SS GRL
>@)HQJ=<X4<DR6HWDO(DUO\SUHGLFWLRQRIGLVHDVH ,&,&&6
SURJUHVVLRQLQ&29,'SQHXPRQLDSDWLHQWVZLWKFKHVW
&7 DQG FOLQLFDO FKDUDFWHULVWLFV 1DW &RPPXQ >@ $ .XVLDN )HDWXUH WUDQVIRUPDWLRQ PHWKRGV LQ GDWD
KWWSVGRLRUJV[ PLQLQJLQ,(((7UDQVDFWLRQVRQ(OHFWURQLFV3DFNDJLQJ
0DQXIDFWXULQJYROQRSS-XO\GRL
>@<DQTLX*H4LQJKXD:DQJ/L:DQJ+RQJKX:X&KHQ
3HQJ -LDMLQJ :DQJ <XDQ ;X *DQJ ;LRQJ <DR\XQ
=KDQJ <LQJSLQJ <L3UHGLFWLQJ ³SRVWVWURNH SQHXPRQLD >@6R]RX3'+DUWVKRUQH*0 7LPHWR3UHJQDQF\$
XVLQJ GHHS QHXUDO QHWZRUN DSSURDFKHV´,QWHUQDWLRQDO &RPSXWDWLRQDO 0HWKRG IRU 8VLQJ WKH 'XUDWLRQ RI 1RQ
-RXUQDO RI 0HGLFDO ,QIRUPDWLFV9ROXPH &RQFHSWLRQIRU3UHGLFWLQJ&RQFHSWLRQ3/2621(
,661 HKWWSVGRLRUJMRXUQDOSRQH
KWWSVGRLRUJMLMPHGLQI >@ : + 2 -DQXDU\ $GROHVFHQW SUHJQDQF\
>@3DXOD'LHKU5REHUW::RRG-DPHV%XVK\KHDG/HLJK $YDLODEOH KWWSVZZZZKRLQWQHZVURRPIDFW
.UXHJHU %DUU\ :ROFRWW 5LFKDUG . 7RPSNLQV VKHHWVGHWDLODGROHVFHQWSUHJQDQF\
³3UHGLFWLRQ RI SQHXPRQLD LQ RXWSDWLHQWV ZLWK DFXWH
924 2021 11th International Conference on Cloud Computing, Data Science & Engineering (Confluence 2021)
Authorized licensed use limited to: Rutgers University. Downloaded on May 17,2021 at 13:58:31 UTC from IEEE Xplore. Restrictions apply.