In e na ional Jou nal o Co e Enginee ing & Managemen
Volume-7, Issue-05, 2023 ISSN No: 2348-9510
367
ENHANCING AI PREDICTIVE ACCURACY THROUGH THIRD-PARTY DATA
INTEGRATION: A CROSS-INDUSTRY RELIABILITY STUDY
Sashi Ki an Vuppala
So wa e De elope ,
I ing, Texas
sashi [email protected]
Abs ac
This esea ch explo es he use o ex e nal da a in AI sys ems, wi h special ocus on da a
eliabili y and o ecas ing p ecision ac oss cons uc ion, e ail, inance, supply chain
managemen , and audi ing sec o s. The esea ch e alua es he in eg a ion o hi d-pa y da a
in o AI modeling echnology and speci ically add esses how his in eg a ion impac s bo h
ope a ional e ec i eness and decision-making p ocesses. The e alua ion o hi d-pa y da a
eliabili y, oge he wi h i s e ec on AI model p edic ion, p o ides e idence ha supe io
quali y da a enhances indus ial ou comes—speci ically in inance and cons uc ion—while
showing limi ed p og ess in e ail, which ope a es unde lowe da a eliabili y s anda ds.
A i icial Neu al Ne wo ks (ANNs) se e as an op imiza ion echnique ha enhances
p edic i e accu acy h ough hei applica ion in e ining model pe o mance. Da a alida ion
amewo ks and go e nance p ocedu es mus be obus ly implemen ed, as he esea ch
iden i ies hese as c i ical componen s in esol ing da a quali y issues. Eme ging echnologies
such as blockchain o e inno a i e ways o gua an ee da a in eg i y. The analysis
demons a es ha hi d-pa y da a in eg a ion is a highly p omising solu ion o AI
applica ions, p o ided da a eliabili y is p ope ly managed o op imize pe o mance.
Keywo ds: A i icial In elligence, Thi d-Pa y Da a In eg a ion, Da a Reliabili y, P edic i e
Accu acy, A i icial Neu al Ne wo ks (ANN), C oss-Indus y Analysis, Blockchain o Da a
In eg i y, Da a Go e nance F amewo ks, Machine Lea ning Op imiza ion, Financial Risk
P edic ion.
I. INTRODUCTION
Mode n echnology demons a es a i icial in elligence sys ems (AI) ac as he undamen al
in eg a i e o ce which combines a ious indus ies a an indus ial scale. Bo h cons uc ion
and e ail oge he wi h inance and audi ing indus ies ha e unde gone signi ican ope a ional
modi ica ions. AI echnology p o ides companies mul iple ools o boos ope a ional success
while deli e ing be e decision making abili ies ha manage secu i y challenges be e . The
pape w i en by Abioye e al. [1] desc ibes how a i icial in elligence echnology a ec s
cons uc ion h ough au oma ed p ocesses and enhances p ojec comple ion a es coupled wi h
In e na ional Jou nal o Co e Enginee ing & Managemen
Volume-7, Issue-05, 2023 ISSN No: 2348-9510
368
be e esou ce dis ibu ion. Technical obs acles om AI implemen a ion need ho ough
heo e ical esea ch o examine da a eliabili y e ec s while p ese ing p edic i e accu acy
le els.
E e y AI sys em ope a es om da a ha unc ions as i s undamen al ope a ional ounda ion.
AI applica ions unc ion e ec i ely when hey possess eliable and dependable ope a ional da a
independen o i s sou ce da abase. P edic i e models hea ily depend on da a eliabili y since
hei ope a ional pe o mance di ec ly esembles he quali y and in eg i y o hei inpu da a.
The main hu dles in AI applica ion o in ech s em om da a quali y alida ion as Giudici [3]
ou lines in his esea ch. The inancial sec o depends on high-quali y da a o p ope isk
managemen because AI-d i en analy ics sys ems assis expe s o measu e c edi isks and
o ecas ma ke beha io o op imize in es men pe o mance. Financial ins i u ions ace
un eliable p edic ions and decisions when hey place oo much ai h in hi d-pa y da a sou ces
since da a e o s and biases along wi h inconsis encies become common.
Bo h big da a analy ics and p edic i e echnologies se e as c ucial ope a ional enhancemen
sys ems h oughou di e en sec o s including audi ing p o essions and supply chain
p ocesses. Huang and Li [4] demons a e ha big da a analysis enhances audi ing quali y
h ough i s capabili y o p o ide audi o s be e quali y in o ma ion o making decisions.
Supplemen ing he accu acy o audi s by enabling isk de ec ion he p edic i e analy ics
unc ions as a subse o big da a. The dependabili y o p edic i e models depends en i ely on
he quali y o da a which hey ecei e o p ocessing. Moses e al. cla i y ha a i icial
in elligence plays a i al ole in audi ing pe o mance because i enables audi o s o base hei
decisions on subs an ial da ase s. Da a usage demons a es why o ganiza ions mus esol e he
p oblems o connec ing hi d-pa y da a o p edic i e models o main aining accu acy and
eliabili y o esul s.
Re ail businesses accep AI echnology as a means o enhance ope a ional e iciency and
cus ome sa is ac ion le els in hei ope a ions. Anica-Popa e al. [2] ecognize ha AI
in eg a ion in e ail ep esen s a p omising g ow h ma ke ha enables imp o ed cus ome
a ge ing and in en o y managemen as well as ailo ed ma ke ing s a egies. The au ho s no e
ha al hough AI b ings a ious bene i s o e ail i encoun e s speci ic da a-quali y p oblems
when implemen ed in his sec o . Companies use ex e nal da a ega ding cus ome ac i i ies
alongside ansac ion in o ma ion when hey make s a egic decisions abou hei businesses.
S a egic decisions based on hese some imes aul y da ase s deli e w ong di ec ions o
o ganiza ions. Anica-Popa e al. [2] s and o a specialized AI concep ual amewo k as a
solu ion o o e come such challenges by gua an eeing da a alidi y and ope a ional u ili y.
AI implemen a ion wi hin inancial ins i u ions c ea es mul iple p oblems because o i s
in eg a ion challenges. The applica ion o AI o inancial isk managemen ecei es analysis
om Al-Blooshi and Nobanee wi hin hei pape [6]. The au ho s p o ide insigh s showing AI
enables inancial o ganiza ions o make be e in es men decisions as well as spo audulen
In e na ional Jou nal o Co e Enginee ing & Managemen
Volume-7, Issue-05, 2023 ISSN No: 2348-9510
369
ac i i ies and deli e be e suppo o clien s. AI sys ems deli e success ul esul s based on he
quali y s anda ds o hei a ailable da a inpu . The p ocess o hi d-pa y da a such as ma ke
ends and economic indica o s unc ions as a key sou ce o AI-based decision-making
acco ding o Al-Blooshi and Nobanee [6]. Such da a collec ion me hods om ex e nal en i ies
ace eliabili y doub s. Financial o ganiza ions isk subs an ial mone a y losses because AI
sys em p edic ions become un eliable when impo an hi d-pa y in o ma ion lacks accu acy
e i ica ion.
The dependency o sus ainable o ganiza ional managemen upon da a eliabili y exis s beyond
p edic i e modeling enhancemen because i main ains business sus ainabili y. The pape by
Jeble e al. [7] explains how supply chain sus ainabili y ini ia i es ecei e suppo om big da a
echnology along wi h p edic i e analy ics. Th ough da a analysis companies achie e be e
supply chain managemen which enables hem o dec ease was e and dec ease hei ca bon
emissions. The in eg a ion o hi d-pa y da a wi hin AI sys ems causes sus ainabili y
di icul ies because o he accu acy p oblems hese sys ems de elop om his da a. The au ho s
in Jeble e al. [7] s ipula e ha ensu ing he quali y and eliabili y o da a ac s as a undamen al
equi emen o eaching sus ainabili y goals h oughou he long e m.
Da a quali y ecogni ion has become c i ical in policy decisions speci ically ela ed o inancial
s abili y. The es ablishmen o da a quali y akes cen e s age in inancial s abili y policy e o s
as explained by Jenkinson and Leono a [8]. The implemen a ion o Legal En i y Iden i ie (LEI)
cons i u es an impo an mo emen owa ds enhancing inancial da a quali y acco ding o hei
pe spec i e. Th ough s anda diza ion o inancial da a epo ing he LEI sys em b ings
inc eased anspa ency oge he wi h diminished isks o inco ec inancial da a. The e o s a e
undamen al o educing dange s ha occu du ing hi d-pa y da a combina ion speci ically in
inancial sys ems whose da a sus ainabili y depends on da a in eg i y.
AI de elopmen akes blockchain echnology as one o i s mos p omising ields o applica ion.
Li e al. [9] demons a e how blockchain echnology enables ci il a ia ion da a secu i y by
p o iding ampe -p oo p o ec ion o a ia ion sys ems. The decen alized ope a ing sys em o
blockchain unc ions as an op imal solu ion o hi d-pa y da a managemen due o i s
unmodi iable and anspa en ansac ion eco ding ea u e. U ilizing blockchain echnology
helps esol e some eliabili y p oblems abou hi d-pa y da a while c ea ing ampe - esis an
e i iable pla o ms. Using blockchain wi h a i icial in elligence sys ems aces di icul ies
mainly because o limi a ions ega ding scalabili y and ollowing egula o y equi emen s.
The in luence o quali y s anda ds ex ends pa icula ly o machine lea ning models when hey
pe o m ac i e lea ning p ocesses. Machine lea ning model pe o mance imp o es along wi h
accu acy h ough ac i e lea ning because i selec s key da a poin s acco ding o Yang e al. [10].
The pe o mance o ac i e lea ning s ongly depends on he excellence s anda ds o inpu da a
in he model. The model exhibi s educed pe o mance when p ocessing aul y o biased
In e na ional Jou nal o Co e Enginee ing & Managemen
Volume-7, Issue-05, 2023 ISSN No: 2348-9510
370
in o ma ion da a ul ima ely leading o w ong p edic ions. The achie emen o ac i e lea ning
elies g ea ly on main aining da a excellence pa icula ly du ing hose pe iods whe e
in o ma ion comes om hi d-pa y sou ces acco ding o Yang e al. [10]. AI sys ems enhanced
by ex e nal da a p o ide undamen al ad an ages o s a egic decision making and ope a ional
e ec i eness and en i onmen al sus ainabili y bene i s h oughou a ious business indus ies.
These da a eliabili y s anda ds unc ion o achie e he success ul implemen a ion o hese
sys ems. Resol ing inaccu acies and biases in AI sys ems is i al o achie e co ec unc ioning
because secu e p edic i e ou pu s depend on such esolu ion. P o essional esea che s mus
wo k wi h p ac i ione s o de elop mode n solu ions which uni e s anda d epo ing sys ems
and blockchain echnology and ac i e lea ning algo i hms o enhancing he p edic i e da a
eliabili y o AI sys ems.
II. REVIEW OF LITERATURE
Many business sec o s use hi d-pa y da a in eg a ion o AI models because he p ac ice
imp o es p edic ion quali y and ope a ional e iciency oge he wi h decision-making abili ies.
The eliabili y o AI sys ems s ems om us ed da a ye ex e nal da a emains c ucial o hei
ope a ion. The p esen e iew deeply e iews exhaus i e esea ch abou ex e nal da a
combina ion o unde s and i s e ec s on eliabili y and o ecas p ecision. This esea ch uses
a ious indus ial sec o s as s udy a eas o analyze p oblems in ex e nal da a acquisi ion
me hods alongside concep ual solu ions o add essing hem.
2.1 A i icial In elligence in he Cons uc ion Indus y
Cons uc ion p o es o be one o many ields which achie e p og ess h ough a i icial
in elligence implemen a ion. The applica ion o AI echnologies in ol ing machine lea ning
(ML) and p edic i e analy ics oge he wi h au oma ion ans o ms he en i e p ojec li ecycle
by op imizing i s di e en s ages. Se e in Abioye e al. [1] es ablishes ha AI ad ances p ojec
planning oge he wi h esou ce managemen and isk assessmen unc ions inside cons uc ion
ac i i ies. Wi h AI sys ems in place i becomes possible o c ea e op imal cons uc ion schedules
which ensu e as ack deli e ies while keeping ma e ials usage a an e icien pace. The
in eg a ion o hi d-pa y da a including wea he o ecas s and ma ke ends in o AI models
equi es a en ion because Abioye e al. [1] s a e his p o es di icul o accomplish. These
p edic ions achie e accu acy le els based on how eliable ex e nal da a sou ces p o e o be
because inconsis encies in hi d-pa y da a may esul in delayed wo k and budge o e uns
and subop imal esou ce deploymen . The impo ance o de eloping a solid amewo k
becomes e iden because i gua an ees high-quali y in eg a ed hi d-pa y da a.
2.2 AI in Re ail: Oppo uni ies and Challenges
AI implemen a ion in he e ail sec o has subs an ially inc eased h ough imp o emen o h ee
co e a eas including cus ome se ice solu ions oge he wi h supply chain con ol along wi h
cus omized ma ke ing s a egies. As pe Anica-Popa e al. [2] e ail business applica ions o
In e na ional Jou nal o Co e Enginee ing & Managemen
Volume-7, Issue-05, 2023 ISSN No: 2348-9510
371
A i icial In elligence empowe i ms o moni o cus ome ac i i y along wi h p oduc
o ecas ing and s ock op imiza ion. AI uses ex e nal consume pu chasing eco ds alongside
demog aphic de ails o boos o ganiza ional decision quali y. Acco ding o Anica-Popa e al. [2]
he in e p e ing o ex e nal da a sou ces p oduces p oblems because i impac s he accu acy and
implemen a ion o da a in eg i y s anda ds. Re ail businesses depend on da a om ex e nal
endo s o unde s and hei cus ome s ye un eliable da a ansmission migh cause hei
s a egies and p oduc alignmen o ail. The p oposed concep ual amewo k om Anica-Popa
e al. [2] es ablishes he necessi y o high-quali y consis en da a in eg a ion as a solu ion o
cu en business challenges.
2.3 The Role o Da a Quali y in Financial Risk Managemen
A i icial in elligence eme ges as he main echnological s anda d in inancial ins i u ions o
imp o ing isk managemen app oaches oge he wi h aud de ec ion and po olio
op imiza ion p ac ices. The pape by Giudici [3] in es iga es how AI g ows in inancial
echnology o enhance p edic i e analy ics which s eng hens decision-making p ocesses. The
in eg a ion p ocess o AI-d i en inancial applica ions depends hea ily on hi d-pa y ma ke
epo s along wi h economic indica o s o hei success. Acco ding o Giudici eliabili y o he
da asou ce s ands as a i al condi ion o each peak pe o mance in AI models. Risk p edic ions
ail because o inco ec o absen in o ma ion in da a sys ems which c ea es subs an ial
inancial ouble o o ganiza ions. The s anda diza ion o ex e nal da a sou ces con inues o be
essen ial o p e en ing in o ma ion e o s ha would diminish AI model pe o mance as
Giudici [3] speci ies. Financial ins i u ions ocus on c ea ing imp o ed da a quali y managemen
echnology alongside es ablishing s ong da a go e nance sys ems o e ec i e challenge
managemen .
2.4 Big Da a and P edic i e Analy ics in Audi ing
Manual audi ing p ocesses ha e expe ienced subs an ial imp o emen because AI and big da a
analy ics echnologies ha e become a ailable. The esea ch pape by Huang and Li [4]
in es iga es big da a u iliza ion o enhance audi pe o ming s anda ds h ough i s in luence
upon aud iden i ica ion and inancial s a emen p ecision. AI sys ems p ocess la ge quan i ies
o in o ma ion o de ec disc epancies which human audi o s would miss in hei audi wo k.
Acco ding o Huang and Li [4] he inco po a ion o hi d-pa y da a ea u ing indus y
benchma ks wi h economic o ecas s in oduces quali y- ela ed obs acles o audi ing sys ems.
Audi esul s will be inaccu a e when in e nal inancial eco ds di e om ma ke in o ma ion.
The eliabili y o hi d-pa y da a mus be ensu ed acco ding o Huang and Li [4] because i
helps audi o s achie e be e p edic ions and a oid inancial miss a emen s.
2.5 AI and Sus ainabili y in Supply Chain Managemen
Supply chain managemen sys ems unde wen signi ican change because o big da a analy ics
coupled wi h a i icial in elligence. The business supply chain ecei es mul iple ad an ages
h ough AI-powe ed p edic i e analy ics sys ems which p ocess ex ensi e da a acco ding o
In e na ional Jou nal o Co e Enginee ing & Managemen
Volume-7, Issue-05, 2023 ISSN No: 2348-9510
372
Jeble e al. [7]. AI sys ems inco po a e p edic i e analy ics which assis s companies in
o ecas ing ma ke equi emen s and maximizes s ock con ol by iden i ying supply chain
ouble spo s. The accu acy o supply chain models p ocessed by AI elies on hi d-pa y da a
inpu s ha co e supplie pe o mance esul s and anspo logis ics s a us acco ding o Jeble
e al. [7]. Ope a ional di icul ies may appea because o da a accu acy p oblems when sys ems
in eg a e ou side in o ma ion sou ces as he au ho s explain. O ganiza ions need o alida e
hei da a while building obus supplie connec ions o sus ain accu a e hi d-pa y
in o ma ion acco ding o Jeble e al. [7].
2.6 Legal En i y Iden i ie (LEI) and Financial Da a Quali y
The inancial sec o main ains ongoing egula o y ac i i ies o enhance he anspa ency along
wi h quali y o da a employed in AI models. The inancial da a quali y enhancemen h ough
Legal En i y Iden i ie (LEI) is analyzed by Jenkinson and Leono a [8]. The LEI sys em se es as
a s anda dized iden i ica ion amewo k o inancial ins i u ions combined wi h ansac ions so
i imp o es bo h da a consis ency and accu acy le els. The in oduc ion o he LEI sys em
ep esen s a key de elopmen o sol e da a quali y p oblems in inancial ins i u ions acco ding
o Jenkinson and Leono a [8]. S anda dized da a o ma s including he LEI when in eg a ed
in o AI models will boos p edic i e accu acy while dec easing he isks ela ed o using
inaccu a e hi d-pa y da a.
2.7 Blockchain o Ensu ing Da a In eg i y in AI Sys ems
The de elopmen o Blockchain echnology p esen s i sel as a possible answe o add ess hi d-
pa y da a in eg a ion p oblems conce ning eliabili y issues. Li e al. [9] demons a e
blockchain implemen a ion in ci il a ia ion by c ea ing an in as uc u e o e i ying o iginal
da a be ween key s akeholde s possesses in eg i y and secu i y. Blockchain echnology
unc ions pe ec ly o hi d-pa y da a managemen due o i s decen alized na u e which
p o ides anspa en ampe -p oo da a s o age o all pa icipa ing pa ies. The au ho s o Li e
al. [9] ecommend ha AI sys ems include blockchain echnology o sol e impo an da a
eliabili y issues speci ically ound in he mos necessa y sec o s. The wide adop ion o
blockchain echnology in AI applica ions equi es esol ing bo h scalabili y and egula o y
issues acco ding o Li e al.[10].
2.8 Ac i e Lea ning o Imp o ing AI Models
The me hod o ac i e lea ning enables AI models o upg ade hei ope a ional pe o mance
h ough he selec ion o aining da a ha con ains he highes in o ma ion alue. The me hods
desc ibed by Yang e al. [10] explain how ac i e lea ning imp o es machine lea ning e iciency
and accu acy h ough e ec i e da a selec ion o ep esen a i e examples. Ac i e lea ning
echniques succeed o he ex en ha he inpu da a emains o high quali y. Reliable da a leads
o be e model ope a ions. Acco ding o Yang e al. [10] he achie emen o ac i e lea ning
echniques depends hea ily on main aining supe io quali y s anda ds o hi d-pa y da a
because subs anda d in o ma ion leads o inadequa e model ou comes.
In e na ional Jou nal o Co e Enginee ing & Managemen
Volume-7, Issue-05, 2023 ISSN No: 2348-9510
373
This e iew es ablishes how hi d-pa y da a inpu s o AI sys ems unc ion ac oss mul iple
sec o s and demons a es ha eliable da a enhances p edic i e capabili ies. Many sec o s such
as cons uc ion, e ail, inance, audi ing, and supply chain managemen ace majo da a quali y
obs acles wi h hei da a con aining w ong in o ma ion and un ai biases along wi h e o s. AI-
d i en sys em success equi es add essing hese obs acles by implemen ing amewo ks along
wi h s anda dized da a o ma s and using eme ging blockchain echnologies. AI e olu ion
equi es addi ional esea ch o disco e no el solu ions which will imp o e bo h da a quali y
s anda ds and p edic i e p ecision o AI models.
III. RESEARCH METHODOLOGY
The esea ch me hod used in his s udy in es iga es how hi d-pa y da a in eg a ion wi h AI
sys ems a ec s p edic i e accu acy and da a eliabili y. The esea ch design oge he wi h he
da a collec ion s a egy and analysis me hods along wi h e alua ion ools o m he basis o his
s udy o es ablish i s eliabili y and alidi y.
A. Resea ch Design
The quan i a i e me hod wi hin his s udy enables esea che s o explo e he connec ion
be ween hi d-pa y da a in eg a ion oge he wi h da a eliabili y and p edic i e accu acy o AI
models. The esea ch implemen a ion has hese key elemen s:
Explo a o y: The s udy examines he e ec s ha inco po a ing ou side da a has on AI
model p edic ion accu acy h oughou a ious business sec o s.
Desc ip i e: Resea ch explo es cu en hi d-pa y da a in eg a ion p ac ices ac oss
sec o s o iden i y p oduc i e ends while assessing di icul ies and bene icial
oppo uni ies ha exis .
Causal-compa a i e: The esea ch p ojec e alua es be ween AI model p edic ion
capabili ies which use ex e nal e sus o ganiza ion-speci ic da a while de eloping
causal ela ionships ha link da a sou ce eliabili y o p edic ed esul s.
B. Popula ion and Sampling
The esea ch a ge s businesses in cons uc ion oge he wi h e ail and inance as well as
audi ing and supply chain managemen who inc easingly use AI wi h p edic i e analy ics.
Expe s opinions and da ase s oge he wi h indus y epo s o med his analy ical sample se .
Pu posi e sampling se es as he esea ch echnique o ga he da a om asso ed indus ies
ha egula ly depend on ex e nal da a sou ces o hei AI modeling needs.
Sample Size: The s udy uses da a om i e indus ies, each wi h a sample size o 50
da ase s collec ed om bo h in e nal and hi d-pa y sou ces.
C. Da a Collec ion
The s udy uses a combina ion o p ima y and seconda y da a collec ion me hods o ga he
In e na ional Jou nal o Co e Enginee ing & Managemen
Volume-7, Issue-05, 2023 ISSN No: 2348-9510
374
ele an in o ma ion:
1. P ima y Da a: This esea ch ga he ed in o ma ion h ough su eys combined wi h
p o essional in e iews om he indus y and da a scien is s and p ac i ione s who wo k wi h
AI. The in e iews analyzed how p o essionals handle ex e nal da a in eg a ion wi hin AI
sys ems while also e i ying p edic ion ou comes o he model.
Su ey Ins umen : The da a collec ion in ol ed he use o a s uc u ed ques ionnai e
which co e ed in o ma ion abou da a quali y along wi h i s sou ces and in eg a ion
issues and hei e ec on p edic i e accu acy.
2. Seconda y Da a: A e iew o indus y epo s alongside jou nal a icles and case s udies
p o ided unde s anding abou cu en p ac ices which combine AI sys ems and hi d-pa y
da a access. The analyzed seconda y da a enabled mo e e ec i e in e p e a ion o da a collec ed
om p ima y sou ces.
Da a Sou ces: Academic jou nals, indus y whi e pape s, and epo s om i ms
in ol ed in AI and da a analy ics.
D. Va iables and Hypo heses
The analysis in es iga es he connec ion be ween da a eliabili y and p edic i e accu acy in AI
models h ough hese impo an a iables:
Independen Va iable: Thi d-pa y da a eliabili y (measu ed by consis ency, accu acy,
comple eness, and imeliness).
Dependen Va iable: P edic i e accu acy o AI models (measu ed by R², MAE, RMSE,
and o he e o me ics).
E. Da a Analysis Techniques
The collec ed da a was analyzed using bo h desc ip i e and in e en ial s a is ics:
1. Desc ip i e S a is ics: Summa y s a is ics (mean, median, s anda d de ia ion) we e used o
cha ac e ize he da a eliabili y and p edic i e accu acy in he i e indus ies. This s ep p o ides
a baseline unde s anding o he ends and pa e ns wi hin he da ase .
2. In e en ial S a is ics: A eg ession analysis was conduc ed o es he ela ionship be ween
da a eliabili y (independen a iable) and p edic i e accu acy (dependen a iable). Mul iple
eg ession models we e used o assess he impac o a ious eliabili y ac o s (e.g., da a
comple eness, accu acy) on AI model p edic ions.
3. A i icial Neu al Ne wo k (ANN) Model: To op imize he p edic i e accu acy o AI models,
an ANN model was employed. The ANN was ained on he da a eliabili y me ics and
p edic i e ou comes, allowing o an ad anced compa ison be ween p edic ed and ac ual
alues.
Model E alua ion: The ANN model’s pe o mance was e alua ed using me ics such as
Mean Squa ed E o (MSE), Roo Mean Squa ed E o (RMSE), and R² sco es.
In e na ional Jou nal o Co e Enginee ing & Managemen
Volume-7, Issue-05, 2023 ISSN No: 2348-9510
375
C oss- alida ion: K- old c oss- alida ion (K=10) was used o p e en o e i ing and
ensu e ha he ANN model gene alized well o unseen da a.
IV. RESULTS DISCUSSION
The s udy explo es he indings ega ding he p ocess o inco po a ing ex e nal da a sou ces
wi h AI sys ems while in es iga ing how da a in eg i y le els a ec p edic i e sys em
pe o mance. The esea che s ob ained da a ega ding AI decision sys ems om di e en
indus ies o his examina ion. The esea ch examines di e en indus ial se ings ha
implemen hi d-pa y da a sou ce in eg a ion in hei AI sys ems whe e cons uc ion joins
e ail alongside inancial domains oge he wi h audi ing p ac ices and supply chain ope a ions.
A e iew o da a eliabili y e ec s on p edic i e accu acy exis s in his esea ch ou come.
A. Da a Reliabili y and P edic i e Accu acy
The main esea ch inqui y examined how in eg a ing ex e nal da a sou ces impac s AI o ecas
accu acy le els. The s udy analyzed he eliabili y o hi d-pa y da a sou ces which di e en
sec o s employ o add ess his issue. The calcula ion o ex e nal da a eliabili y ocused on h ee
aspec s: consis ency, accu acy, and comple eness alongside imeliness.
Table 1: Da a Reliabili y Index o Thi d-Pa y Da a Ac oss Indus ies
Indus y
Consis ency
(%)
Accu acy
(%)
Comple eness
(%)
Da a
Reliabili y
Index (%)
Cons uc ion
85
88
80
84
Re ail
75
70
68
71
Finance
90
92
89
90
Audi ing
82
85
81
82
Supply
Chain
78
77
76
77
In e na ional Jou nal o Co e Enginee ing & Managemen
Volume-7, Issue-05, 2023 ISSN No: 2348-9510
382
wi hin bo h AI models and ex e nal cus ome da a in eg a ion. Th oughou e e y indus y he
ANN model showcased success ul pe o mance by demons a ing minimal a ia ions be ween
o ecas ed esul s and ac ual measu emen s. The analysis esul s demons a e ha ANN
unc ions success ully o p edic bo h p edic i e accu acy R² and he MAE and RMSE alues in
hi d-pa y da a-suppo ed AI models. Thus he e idence implies ANN p o ides impo an
op imiza ion capabili ies o AI models speci ically when p ocessing ex e nal da a. The
eliabili y and p ecision o p edic ions become in eg al because ANN-based op imiza ion
p o es essen ial o ensu e hi d-pa y da a in eg a ion main ains i s ideli y wi h p edic ions.
The model equi es an addi ional imp o emen s ep which includes es ing mo e complex
a chi ec u al designs and adding ex a ea u es o maximize p edic ion accu acy.
V. CONCLUSION
The esea ch analyzes how AI applica ions use hi d-pa y da a p ocessing impac s da a
c edibili y and p edic i e s eng h in business ela ionships be ween he cons uc ion sec o and
e ail- inance o ganiza ions and audi p ac ices as well as supply chain applica ions. The
p edic i e powe o AI models depends speci ically on he accu acy s anda ds and
measu emen quali y o in o ma ion p o ided by ex e nal endo s. The p edic i e abili ies o
cons uc ion businesses and inancing o ganiza ions s eng hened because hey used high-
quali y da a om ex e nal endo s. Limi ed p og ess was obse ed in he e ail indus y
because he exis ing da a quali y issues wi hin ope a ions p e en ed e ec i e esul s. A i icial
Neu al Ne wo ks (ANN) op imiza ion echniques p oduced be e p edic ion ou comes because
hey demons a e ha ANN echnology p o ides supe io pe o mance o in eg a ed da a
p ocessing in AI models. This esea ch shows se e al quali y- ela ed issues ha p o e he
necessi y o de eloping obus da a p o ec ion go e nance amewo ks wi h da a alida ion
sys ems. Mode n blockchain inno a ions explo e how sys ems can gain enhanced da a
anspa ency alongside p o ec ed da a in eg i y. AI sys ems bene i g ea ly om quick and
dependable hi d-pa y da a in eg a ion when eliabili y ools enhance he end esul s. The ield
equi es addi ional esea ch o imp o e da a quali y managemen sys ems along wi h complex
model sys ems ha handle hi d-pa y da a complexi ies.
REFERENCES
1. S. O. Abioye, L. O. Oyedele, L. Akanbi, A. Ajayi, J. M. D. Delgado, M. Bilal, and A.
Ahmed, "A i icial in elligence in he cons uc ion indus y: A e iew o p esen s a us,
oppo uni ies and u u e challenges," Jou nal o Building Enginee ing, ol. 44, p. 103299,
2021.
2. I. Anica-Popa, L. Anica-Popa, C. Rădulescu, and M. V încianu, "The in eg a ion o
a i icial in elligence in e ail: bene i s, challenges and a dedica ed concep ual
amewo k," Am i ea u Economic, ol. 23, no. 56, pp. 120-136, 2021.
In e na ional Jou nal o Co e Enginee ing & Managemen
Volume-7, Issue-05, 2023 ISSN No: 2348-9510
383
3. P. Giudici, "Fin ech Risk Managemen : A Resea ch Challenge o A i icial In elligence in
Finance," F on ie s in A i icial In elligence, ol. 1, no. 1, 2018.
4. H. Huang and S. Li, "Can Big Da a analysis and p edic i e analy ics imp o e audi ing
quali y?" Audi ing: A Jou nal o P ac ice & Theo y, ol. 36, no. 4, pp. 21-45, 2017.
5. O. D. Moses, S. Nwankwo, and A. Oyedijo, "A i icial In elligence and Audi o s'
Pe o mance in Nige ia: Implica ions o Enhanced Audi Quali y," Jou nal o
Accoun ing, Business and Finance Resea ch, ol. 4, no. 1, pp. 58-65, 2018.
6. L. Al-Blooshi and H. Nobanee, "Applica ions o a i icial in elligence in inancial
managemen decisions: a mini- e iew," SSRN, no. 3540140, 2020. [Online]. A ailable:
h ps://dx.doi.o g/10.2139/ss n.3540140
7. S. Jeble, R. Dubey, S. J. Childe, T. Papadopoulos, D. Roubaud, and A. P akash, "Impac
o big da a and p edic i e analy ics capabili y on supply chain sus ainabili y," The
In e na ional Jou nal o Logis ics Managemen , ol. 29, no. 2, pp. 513-538, 2018.
8. N. Jenkinson and I. S. Leono a, "The impo ance o da a quali y o e ec i e inancial
s abili y policies-Legal en i y iden i ie : a i s s ep owa ds necessa y inancial da a
e o ms," Financial S abili y Re iew, ol. 17, pp. 101-110, 2013.
9. J. Li, Z. Peng, A. Liu, L. He, and Y. Zhang, "Analysis and u u e challenge o blockchain
in ci il a ia ion applica ion," in P oc. 2020 IEEE 6 h In e na ional Con e ence on
Compu e and Communica ions (ICCC), Chengdu, China, 11–14 Dec. 2020, pp. 1742–
1748.
10. Y. Yang, M. Loog, and T. M. Hospedales, "Ac i e lea ning by que ying in o ma i e and
ep esen a i e examples," IEEE T ans. Pa e n Anal. Mach. In ell., ol. 40, no. 10, pp.
2436–2450, 2018.
