Credit card fraud detection through machine learning

credit card fraud detection using machine learning

1. Credit Card Fraud Detection using Machine
learning
Data Alcott Systems
Ph/ Whatsapp [phone]
Check demo: www.finalsemprojects.com
Data Alcott Systems Contact 9600095046

2. Duetoincreaseoffraudwhichresultsinlossofmoneyacrosstheglobe,several
methodologiesandtechniquesdevelopedfordetectingfraudsFrauddetectioninvolves
analysingtheactivitiesofusersinordertounderstandthemaliciousbehaviourofusers.
Maliciousbehaviourisabroadtermincludingdelinquency,fraud,intrusion,andaccount
defaulting.Thispaperpresentsasurveyofcurrenttechniquesusedincreditcardfraud
detectionandevaluatesanewhybridapproachtoidentifyfrauddetection.Intheproposed
work,weanalyzecreditcardfrauddetectionusingmachinelearningalgorithmnamelylogistic
regressionandDecisionTree.Tomakethelearningprocessefficient,weusedPrincipal
componentforfeatureselection.
Abstract
Data Alcott Systems Contact 9600095046

3. Algorithm Used
Principal component analysis
Feature selection
Decision Tree and Regression
Decision Tree and Regression used for fraud detection
Data Alcott Systems Contact 9600095046

4. Introduction
Withtheemergingriseoftechnologytoday,thedependencyone-commerceandtheonline
paymentshasgrownexponentially.Asthecreditcardprovidesconveniencetotheusersbutfrauds
causedduetotheseactivitiescausesinconvenience.Thecreditcardinformationisconfidential,
thebankandtheotherfinancialenterprisesdoesn'twanttodisclosetheinformationabouttheir
customers.Riskmanagementiscriticalforfinancialenterprisestosurviveinsuchcompeting
industry.Theprovisionallossarisesduetothe“bad”accountsbanklendsthemoneytocustomers
whoeventuallydonothavecapabilitytopayback.Intheriskmanagement,thechancesoffalse
negative(false“good”accounts)couldstillbehigh.However,byleveragingtheirperformancesuch
ascreditcardutilization,paymentinformation,riskscanfurtherbemanagedtocontrol
provisionalloss.Inthispaper,afocusonriskmanagementaswellasfrauddetectionisdepicted.
Data Alcott Systems Contact 9600095046

5. Figure: Overview of Machine learning
Data Alcott Systems Contact 9600095046

6. Existing System
AdvancedoversamplingmethodslikeSMOTEgeneratesynthetictraininginstancesfrom
theminorityclassbyinterpolation,insteadofsamplereplication.
TheweightingstrategyofAdaBoostisequivalenttoresamplingthedataspace[6],which
areapplicabletomostclassificationsystemswithoutchangingtheirlearningmethods.
Bahnesonet.alinexpandedthetransactionaggregationstrategy,andproposedtocreate
anewsetoffeaturesbasedonanalyzingtheperiodicbehaviorofthetimeofa
transactionusingthevonMisesdistribution.Then,usingarealcreditcardfrauddataset
providedbyalargeEuropeancardprocessingcompany.
Halvaieeet.alinaddressedcreditcardfrauddetectionusingArtificialImmuneSystems
(AIS),andintroducedanewmodelcalledAIS-basedFraudDetectionModel(AFDM).
Theauthorusedanimmunesysteminspiredalgorithm(AIRS)andimproveditforfraud
detection.
Data Alcott Systems Contact 9600095046

7. Disadvantages
AISoptimizedthedataandlearningratebuttheaccuracyarrivedisless.
Differentsetsoffeatureshadanimpactontheresults.
SOMTEdrawbacksisthatundersamplingmaylosesomepotential
information,andoversamplingmayleadtheoverfitting.
AdaBoostismostlyusedclassificationwhereasextralearningcostisaburden.
Data Alcott Systems Contact 9600095046

8. Proposed System
Creditcardlogsminingisoneofthemostsignificantfieldsinthe
areaofdatamining.Therehavebeenalargenumberofdatamining
algorithmsrootedinthesefieldstoperformdifferentdataanalysis
tasks.
ApplyPrincipalcomponentanalysisalgorithmtofindbestfeatures,
thenapplydecisiontreeclassifierforcreditcardfraudprediction.
Data Alcott Systems Contact 9600095046

9. Advantages
Bestaccuracyforthestudy
PCAgetsbestfeatures
Data Alcott Systems Contact 9600095046

10. SYSTEM REQUIREMENTS
Software Requirements
1.Windows Xp, Windows 7(ultimate, enterprise)
2.Python 3.6 and related libraries
Hardware Components
1.Processor –i3
2.Hard Disk –5 GB
3.Memory –1GB RAM
Data Alcott Systems Contact 9600095046

11. SYSTEM ARCHITECTURE
Data Alcott Systems Contact 9600095046

12. IMPLEMENTATION
TheproposedworkisimplementedinPython3.6.4withlibrariesscikit-
learn,pandas,matplotlibandothermandatorylibraries.Wedownloaded
datasetfromkaggle.com.Thedatadownloadedcontainstrainsetandtest
setseparatelywithtwodifferentclasses0and1.Thetraindataset
consideredastrainsetandtestdatasetconsideredastestset.Machine
learningalgorithmisappliedsuchaslogisticregressionandDTused.
Data Alcott Systems Contact 9600095046

13. DATA PRE-PROCESSING
Wehavetakenmultipleattributeinourcasestudy,dataset16features/
attributesaretakenfrstudy.Pre-processingofdatasetisdonefor
convertingthestringattributestonumeralsandmissingdatarecordsare
dropped.Thepre-processeddataisstoredin“dataset.csv”file,whichis
givenasinputformachinelearningmodels.
Data Alcott Systems Contact 9600095046

14. Principal component analysis feature selection
PrincipalComponentAnalysis,orPCA,isadimensionality-reduction
methodthatisoftenusedtoreducethedimensionalityoflargedatasets,
bytransformingalargesetofvariablesintoasmalleronethatstill
containsmostoftheinformationinthelargeset.
Reducingthenumberofvariablesofadatasetnaturallycomesatthe
expenseofaccuracy,butthetrickindimensionalityreductionistotradea
littleaccuracyforsimplicity.Becausesmallerdatasetsareeasiertoexplore
andvisualizeandmakeanalyzingdatamucheasierandfasterformachine
learningalgorithmswithoutextraneousvariablestoprocess.
Data Alcott Systems Contact 9600095046

15. EXPERIMENTAL RESULTS AND EVALUATIONS
Implementedtwomachinelearningalgorithmonthegivendatasetfor
creditcardfrauddetectionshowsthatLogisticsregressionmodel
outperformsothermodels.TheaccuracyofRegressionishighcompared
todecisiontreeclassificationmachinelearningalgorithms.
Algorithm Accuracy
DecisionTree 71.41
LogisticRegression79.91
Data Alcott Systems Contact 9600095046

16. Metrics
Data Alcott Systems Contact 9600095046

17. CONCLUSION
Inthisstudy,anewmethodfordatagenerationofimbalanceddataset's
minorityclasswasproposedtoenhancefrauddetectionincreditcardby
machinelearningandPCAalgorithmasanoversamplingstrategy.
AlthoughPCAalgorithmshavebeenappliedinmanyareas,our
applicationdomainaimstohandleimbalanceddatasetissueby
generatingnewminorityclassinstancestogainnewtrainingsets.
Applyingthisalgorithmintobankcreditcardfrauddetectionsystemaims
toreducefraudulenttransactionanddecreasethenumberoffalsealert.
Data Alcott Systems Contact 9600095046

18. FUTURE ENHANCEMENTS
Afurtherworkistoimplementthisapproachusingpythonprogramming
language,thiswillallowustovalidateourworkandproducepertinent
experimentalresults.
Data Alcott Systems Contact 9600095046

19. REFERENCES
[1] H. Lei et al., ``A deeply supervised residual network for HEp-2 cell classification via cross-modal transfer learning,'' Pattern Recognit.,
vol. 79, pp. 290 302, Jul. 2018.
[2] P. Wang, L. Li, Y. Jin, and G. Wang, ``Detection of unwanted traffic congestion based on existing surveillance system using in freeway
via a CNN-architecture trafficNet,'' in Proc. 13th IEEE Conf. Ind. Electron. Appl., May/Jun. 2018, pp. 1134 1139.
[3] X. Zhu, Y.Wang, J. Dai, L. Yuan, and Y.Wei, ``Flow-guided feature aggregation for video object detection,'' in Proc. ICCV, Mar. 2017, pp.
408 417.
[4] Z. Zhao,W. Chen, X.Wu, P. C. Chen, and J. Liu, ``LSTM network: A deep learning approach for short-term traffic forecast,'' IET Intell.
Transp. Syst., vol. 11, no. 2, pp. 68 75, Mar. 2017.
[5] P. Li, D. Wang, L. Wang, and H. Lu, ``Deep visual tracking: Review and experimental comparison,'' Pattern Recognit., vol. 76, pp.
323 338, Apr. 2018.
[6] P. Wang and J. Di, ``Deep learning-based object classification through multimode fiber via a CNN-architecture SpeckleNet,'' Appl.
Opt., vol. 57, no. 28, pp. 8258 8263, 2018.
[7] J. Zhao, Z. Zhang, W. Yu, and T.-K. Truong, ``A cascade coupled convolutionalneural network guided visual attention method for ship
detection from SAR images,'' IEEE Access, vol. 6, pp. 50693-50708, 2018.
[8] T. Pamula, ``Road traffic conditions classification based on multilevel filtering of image content using convolutionalneural
networks,'' IEEE Intell. Transp. Syst. Mag., vol. 10, no. 3, pp. 11 21, Jun. 2018.
[9] M. Barth and K. Boriboonsomsin, ``Environmentally beneficial intelligent transportation systems,'' IFAC Proc. Volumes, vol. 42, no.
15, pp. 342 345, 2009.
[10] A. Krizhevsky, I. Sutskever, and G. E. Hinton, ``Imagenetclassification with deep convolutionalneural networks,'' in Proc. Adv.
Neural Inf. Pro-cess. Syst., 2012, pp. 1097 1105.
Data Alcott Systems Contact 9600095046

20. Thank You
Data Alcott Systems Contact 9600095046