Novel Financial Applications of Machine Learning and Deep Learning: Algorithms, Product Modeling, and Applications PDF
Preview Novel Financial Applications of Machine Learning and Deep Learning: Algorithms, Product Modeling, and Applications
International Series in Operations Research & Management Science Mohammad Zoynul Abedin Petr Hajek Editors Novel Financial Applications of Machine Learning and Deep Learning Algorithms, Product Modeling, and Applications International Series in Operations Research & Management Science FoundingEditor FrederickS.Hillier,StanfordUniversity,Stanford,CA,USA Volume 336 SeriesEditor Camille C. Price, Department of Computer Science, Stephen F. Austin State Uni- versity,Nacogdoches,TX,USA EditorialBoardMembers Emanuele Borgonovo, Department of Decision Sciences, Bocconi University, Milan,Italy BarryL.Nelson,DepartmentofIndustrialEngineering&ManagementSciences, NorthwesternUniversity,Evanston,IL,USA BruceW.Patty,VeritecSolutions,MillValley,CA,USA MichaelPinedo,SternSchoolofBusiness,NewYorkUniversity,NewYork,NY, USA RobertJ.Vanderbei,PrincetonUniversity,Princeton,NJ,USA AssociateEditor Joe Zhu, Foisie Business School, Worcester Polytechnic Institute, Worcester, MA, USA The book series International Series in Operations Researchand Management Science encompasses the various areas of operations research and management science. Both theoretical and applied books are included. It describes current advances anywhere intheworld thatareatthecuttingedgeofthefield.Theseries is aimed especially at researchers, advanced graduate students, and sophisticated practitioners. Theseriesfeaturesthreetypesofbooks: (cid:129)Advancedexpositorybooksthatextendandunifyourunderstandingofpartic- ularareas. (cid:129)Researchmonographsthatmakesubstantialcontributionstoknowledge. (cid:129) Handbooks that define the new state of the art in particular areas. Each handbook will be edited by a leading authority in the area who will organize a team of experts on various aspects of the topic to write individual chapters. A handbook may emphasize expository surveys or completely new advances (either researchorapplications)oracombinationofboth. Theseriesemphasizesthefollowingfourareas: MathematicalProgramming: Including linearprogramming, integerprogram- ming, nonlinear programming, interior point methods, game theory, network opti- mization models, combinatorics, equilibrium programming, complementarity theory, multiobjective optimization, dynamic programming, stochastic program- ming,complexitytheory,etc. Applied Probability: Including queuing theory, simulation, renewal theory, Brownian motion and diffusion processes, decision analysis, Markov decision processes, reliability theory, forecasting, other stochastic processes motivated by applications,etc. ProductionandOperationsManagement:Includinginventorytheory,produc- tion scheduling, capacity planning, facility location, supply chain management, distributionsystems,materialsrequirementsplanning,just-in-timesystems,flexible manufacturing systems, design of production lines, logistical planning, strategic issues,etc. Applications of Operations Research and Management Science: Including telecommunications,healthcare,capitalbudgetingandfinance,economics,market- ing, public policy, military operations research, humanitarian relief and disaster mitigation,serviceoperations,transportationsystems,etc. ThisbookseriesisindexedinScopus. (cid:129) Mohammad Zoynul Abedin Petr Hajek Editors Novel Financial Applications of Machine Learning and Deep Learning Algorithms, Product Modeling, and Applications Editors MohammadZoynulAbedin PetrHajek DepartmentofFinance,Performanceand FacultyofEconomicsandAdministration Marketing UniversityofPardubice TeessideUniversityInternationalBusiness Pardubice,CzechRepublic School,TeessideUniversity Middlesbrough,UK ISSN0884-8289 ISSN2214-7934 (electronic) InternationalSeriesinOperationsResearch&ManagementScience ISBN978-3-031-18551-9 ISBN978-3-031-18552-6 (eBook) https://doi.org/10.1007/978-3-031-18552-6 ©TheEditor(s)(ifapplicable)andTheAuthor(s),underexclusivelicensetoSpringerNatureSwitzerland AG2023 Thisworkissubjecttocopyright.AllrightsaresolelyandexclusivelylicensedbythePublisher,whether thewholeorpartofthematerialisconcerned,specificallytherightsoftranslation,reprinting,reuseof illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similarordissimilarmethodologynowknownorhereafterdeveloped. Theuseofgeneraldescriptivenames,registerednames,trademarks,servicemarks,etc.inthispublication doesnotimply,evenintheabsenceofaspecificstatement,thatsuchnamesareexemptfromtherelevant protectivelawsandregulationsandthereforefreeforgeneraluse. The publisher, the authors, and the editorsare safeto assume that the adviceand informationin this bookarebelievedtobetrueandaccurateatthedateofpublication.Neitherthepublishernortheauthorsor theeditorsgiveawarranty,expressedorimplied,withrespecttothematerialcontainedhereinorforany errorsoromissionsthatmayhavebeenmade.Thepublisherremainsneutralwithregardtojurisdictional claimsinpublishedmapsandinstitutionalaffiliations. ThisSpringerimprintispublishedbytheregisteredcompanySpringerNatureSwitzerlandAG Theregisteredcompanyaddressis:Gewerbestrasse11,6330Cham,Switzerland Preface TheNovelFinancialApplicationsofMachineLearningandDeepLearning:Algo- rithms, Product Modelling, and Applications presents the state of the art of the application of machine learning (ML) and deep learning (DL) in the domain of finance. We will present a combination of empirical evidence to diverse fields of financesothatthisbookisusefultoacademics,practitioners,andpolicymakerswho arelookingtotrainnovelandthemostadvancedmachinelearningclassifiers.Thus, the purpose of this book is to provide a broad area of applications to different financial assets and markets. Furthermore, from an extensive literature assessment, it is evident that there are no existing textbooks that narrate ML and DL to unlike areasoffinanceortoanextensiverangeofproductsandmarkets. Businessrisk and uncertaintycertainly arethetoughestchallengeinthe finance domainfacedbymanyresearchersandmanagers.Suchuncertaintytherebyinitiates an unavoidable risk factor, which is a fundamental element of financial theory. To the best of our knowledge, the financial domain has not been a focused subject- matterforgoodMLrelatedbooks.Thereisalsoascarcityofinformationabouthow financial enterprises supervise crisis events and achieve turnaround. In order to fix the multifarious nature of the financial problem, this edited book advocates inter- disciplinaryapproachesbasedonmachinelearning. Machine learning is involved in the analysis of large and multiple feature instances. It principally refers to acquiring knowledge and intelligence (by a com- puterprogram)fromaprocessedtrainingexampleforgeneratingpredictions.Itdeals with computationallyintensivetechniques,suchascluster analysis, dimensionality reduction,andsupportvectoranalysis.Itisprincipallytheareaofcomputerscience andisalreadyfrequentlyappliedinsocialsciences,financeandbanking,marketing research, operations research, and applied sciences. Moreover, computational finance is a domain of applied computer science that is concerned with practical issues in finance. It may be characterized as the study of features, instances, and learning algorithms applied in finance. It is an interdisciplinary area that integrates computational tools with numerical finance. Furthermore, computational finance applies arithmetical proofs that can be fitted to economic experiments, thereby v vi Preface contributingtotheadvancementoffinancialdatamodelingtechniquesandsystems. Thesecomputationaltechniquesareutilizedinfinancialriskmanagement,corporate bankruptcy prediction, stock price prediction, and portfolio management. Finally, thisproposedtextbookcouldplayanimportantroleinfinancialdatalearning. Besides, this volume will be a basis for empirical and theoretical practices. The empirical experiments aim to minimize financial risk and uncertainty by covering andfittingthemostadvancedandnovelmachinelearningalgorithms.Moreover,it generates academic literature as well as financial product and finance modeling inferencestowardcustomercreditriskassessment,datamining,patternrecognition, bankruptcyprediction,andsoon.Tobespecific,thevolumeisbroadlydividedinto three parts, with the first set of chapters focusing on the recent trend and issues of financial technology (FinTech). The second set of chapters comprises empirical essays on the prediction and forecasting financial risk by applying ML and DL tools and techniques. The third set of chapters combines empirical evidence of financial time-series data forecasting. The volume ends with a set of emerging technologiesinfinancialeducationandhealthcareandtheirempiricalapplications. Part 1: Recent Developments in FinTech ThefirstpartpresentsfourchaptersonrecentdevelopmentinFinTech. Chapter “FinTech Risk Management and Monitoring” focuses on risk manage- ment and monitoring in FinTech. The recent emergence of financial technology innovationsinthefinancialservicesandsomesignificantrisksareinvestigatedusing the qualitative research method. Additionally, the appropriate way to mitigate the risk is discussed in this chapter. Besides this objective, this chapter discusses the major risk behind the rapid development of fintech and the steps for fintech risk management.Thefourkeyregulatorytechniquesthathaveimportantapplicationsin FinTech management and monitoring are added, and, finally, the chapter summa- rizesthemainchallengesofFinTechriskmanagement. Chapter “Digital Transformation of Supply Chain with Supportive Culture in BlockchainEnvironment”explorestheinfluenceofblockchainonthedigitaltrans- formation of Supply Chain Management (SCM). This chapter is also aimed to determine the importance of supportive culture in the adoption of blockchain in supply chains. The study findings indicate that the digitalization of supply chain managementbyadoptingblockchaintechnologyispositivelycorrelatedwithorga- nizationalprosperity.Thechapteralsoindicatesthatsupportivecultureiscrucialto practicingblockchaintechnology.Thisstudysuggeststhatpolicymakersandstake- holders ensure a supportive culture to establish a traceable, efficient, and effective supplychain. Chapter “Integration of Artificial Intelligence Technology in Management Accounting Information System: An Empirical Study” conducts an empirical studyontheintegrationofartificialintelligencetechnologyinmanagementaccount- ing information systems. This study established an artificial neural network-based Preface vii model to predict management information and verify the accuracy of the model using some real data. Five dimensions are considered to develop the model, accounting analysis management system, accountingdecision supportsystem, per- formance management information system, risk management information system, andenvironmentmanagementinformationsystem. The essentiality to analyze big data in accounting and finance is discussed in Chap.“TheImpactofBigDataonAccountingPractices:EmpiricalEvidencefrom Africa”. Evidence indicates that big data significantly impact accounting and auditing accounting, utilizing the diversity of data volume, data variety, and data velocity. Chapter “The Impact of Big Data on Accounting Practices: Empirical Evidence from Africa” shows the impact of big data on accounting practices, and thestudyareaisAfrica.Themaingoalofthischapteristoexploretheimpactsofbig data on accounting using accountants in Nigeria. Multiple regression is used for 151responses, andsamplesarecollectedusingtherandomsamplingmethod.This study proves that big data positively and significantly affect financial reporting, performance measurement, corporate budgeting, audit evidence, risk management, andfraudmanagement.Thisstudyhelpsaccountants,prospectiveaccountants,and accountinggraduatesintheirstudies. Part 2: Financial Risk Prediction Using Machine Learning ThesecondpartcontainsfourchaptersthatdiscusstheapplicationsofMLandDL approachestopredictandforecastfinancialrisk. Chapter “Using Outlier Modification Rule for Improvement of the Performance of Classification Algorithms in the Case of Financial Data” discusses how to improve classifier performance by mining and modifying outliers of financial datasets. This chapter offers insights into the Financial Decision Support System for financial decision makers. This study employs four distinct classification algo- rithms such as linear discriminant analysis, k-nearest neighbor, naïve Bayes, and supportvectormachineforbothoriginalandmodifieddatasetstodetectcreditcard fraud. The study’s findings show that the classifiers perform better on modified datasetsthanonoriginalcreditcarddatasets. Chapter “Default Risk Prediction Based on Support Vector Machine and Logit SupportVectorMachine”isapredictiveanalysisofthemachinelearningalgorithm fordefaultriskprediction.ThisstudyproposesaLogitSVM modelthathybridized thetraditionalsupportvectormachinewithpopularlogisticregressiontoassessthe credit default risk. The authors use real-world credit databases to validate the probability and value of the proposed model. Type I error, type II error, and root mean square error (RMSE) are used to evaluate the performance of the regressors. Empirical findings show that the proposed hybrid model is superior to maximize accuracy and minimize RMSE. This chapter helps stockholders develop a wide varietyofapproachestopredictthecreditcustomers’defaultrisk. viii Preface Chapter “Predicting Corporate Failure Using Ensemble Extreme Learning Machine” shows the corporate failure prediction using the Ensemble Extreme Learning Machine. The claim is that the early-stage prediction of corporate failure is essential for banks and financial institutions to solve financial decision-making problems. Newly developed artificial intelligence technique Extreme Learning Machine has an extremely fast learning classifier. To prove the superiority of this method, the authors compare the result with four benchmark ensemble methods, namelymultipleclassifiers,bagging,boosting,andrandomsubspace.Experimental resultsonFrenchfirmsindicatedthatbaggedandboostedextremelearningmachines showedthebest-improvedperformance. Chapter “Assessing and Predicting Small Enterprises’ Credit Ratings: A Multicriteria Approach” focuses on small enterprises; it assigns and predicts the small enterprise’s credit rating using a multicriteria approach. In reality, small enterprises have made it difficult for financial institutions such as commercial banks to accurately determine the credit risk, creating salient loan difficulties due toshorttime,highfrequency,urgentdemandforcredit,andasmallnumberoftheir loans.Tosolvethisissue,thechapterdevelopsanewapproachforassessingcredit riskinsmallenterprisesbycombininghigh-dimensionalattributereductionmethods withfuzzyC-meanstogradethecreditratingsofenterprisesrequestingloans. Part 3: Financial Time-Series Forecasting The third part contains two chapters that explore empirical evidence of time-series datamodeling. Chapter“AnEnsembleLGBM(LightGradientBoostingMachine)Approachfor Crude Oil Price Prediction” is on the prediction of crude oil prices. Every second countswhengovernments,businesses,andindividualsneedtoknowwhatthefuture of thecrude oil marketwill bring in terms ofpricing.Estimating thefuture cost of crude oil is a crucial step toward building an economy that can last. In order to effectivelypredictfuturecrudemarketprices,thisresearchwillusemachinelearning andensemblelearningtechniques.Themodelusinglightgradientboosting(LGBM) is proposed by the authors to predict the price of crude oil. By analyzing and modeling the Brent time-series crude oil data, the accuracy and precision of our predictorscanbeimproved.TheLBGMforecastiscomparedtothelassoregression, randomforestregression,anddecisiontreeregressionmethods.Theresultsachieved by the suggested model are quite similar to and better than those obtained by the baseline model when measured using RMSE, mean absolute percentage error (MAPE),meansquarederror(MSE),andmeanabsoluteerror(MAE). Chapter “Model Development for Predicting the Crude Oil Price: Comparative EvaluationofEnsembleandMachineLearningMethods”alsoshowstheprediction ofcrudeoilpricesusingdifferentmethods.Thisstudyshowsacomparativestudyof ensemble algorithms and machine learning algorithms to find the best forecasting model.Thisresearch uses machine learning and an ensemblealgorithm to forecast Preface ix crudeoilprices,anditcomparestheefficacyofthreedifferentregressionmodels— AdaBoost, Bagging Lasso, and Support Vector Regression—to conclude which is the most suitable. Time-series data on crude oil prices are analyzed and used to validate the forecasting model. The results of the various algorithms are compared using an actual vs. anticipated curve. According to the results, the ensemble AdaBoostmethodhassuperiorperformance.Themeansquareerror,meanabsolute error,rootmeansquareerror,meanabsolutepercentageerror,variancescore,andR2 areusedtoverifytheoutcome.Thisresearchwillhelpthosewithastakeinthecrude oilindustrydecideandcraftpoliciesbasedonprojectedfutureprices. Part 4: Emerging Technologies in Financial Education and Healthcare The fourth part contains three chapters that explore the financial education and healthcareissuesandtheiremergingtrends. Chapter “Discovering the Role of M-Learning Among Finance Students: The Future of Online Education” investigates the role of m-learning among finance studentsandthefutureofonlinehighereducation.Thisstudyaimstofindthehidden issuesofm-learninginfinancestudies.Thisstudyismainlyaqualitativeapproach, andthefindingsshowthatdigitalizededucationprovidestheopportunityformajor financestudentstoaccessfinancialmarketsusingtheInternetandgainpersonaland professionalknowledgeinabetterwayratherthantraditionallearning.Italsoshows thatm-learninghasasignificantpositiverelationshipwiththeeffectivenessofonline education.Thisanalysishasasignificantimplicationforeducationpolicymakersand practitioners. Chapter “Exploring the Role of Mobile Technologies inHigher Education: The Impactof Online Teaching on Traditional Learning” demonstrates how technolog- ical evolvements derive the conduction of higher education, especially mobile technology. This study also intended to detect the factors that attract pupils who donotadoptanonlineeducationsystem.Aqualitativeapproachisusedtodetermine the pros and cons of the technology-based education system in universities. The authors reveal that the adoption of mobile technologies in academic education enablesstudentstoaccessvaluableresourcesfreeofcostandeffortlessly,whichin turn helps them to develop strong knowledge and understanding of their study contents. This study opens up a new arena for research scholars to discover the importanceofonlineeducationsystems. Chapter“KnowledgeMiningfromHealthData:ApplicationofFeatureSelection Approaches”assessedtheperformanceoffeatureselectiontechniquesinknowledge mining of health datasets. This study compared seven popular knowledge mining approaches on six popular Affymetrix and cDNA datasets. Employing a support vectormachineclassifier,thestudydeterminedtheknowledgeminers’accuracyand areaunderthecurvevalues.Thefindingofthischapterinformsthatthesimplelasso
The list of books you might like
