Final Exam Project Layout

docx

School

Mississippi College *

*We aren’t endorsed by this school

Course

640

Subject

Information Systems

Date

Apr 3, 2024

Type

docx

Pages

7

Uploaded by mmakol

Report
Running Head: Final Exam Project Layout UNIVERSITY OF THE POTOMAC DACS640:7: Online Data Integration, Warehousing, Provenance, and Analysis Dr. Daryl R. Brydie Megha Makol, Quoc Bao Tran, Cassidy Ballard, Roshni Patel
Final Exam Project Layout Abstract The goal of this project is to detect fraudulent credit card transactions using machine learning techniques to prevent fraudsters from using customers' accounts in an unauthorized manner. Credit card fraud is increasing rapidly around the world, which is why action should be taken to stop fraudsters. Putting a limit on those actions would benefit customers because their money would be recovered and retrieved back into their accounts, and they would not be charged for items or services that they did not purchase, which is the main goal of the project. The goal of this project is to detect fraudulent credit card transactions using machine learning techniques to prevent fraudsters from using customers' accounts in an unauthorized manner. Credit card fraud is increasing rapidly around the world, which is why action should be taken to stop fraudsters. Putting a limit on those actions would benefit customers because their money would be recovered and retrieved back into their accounts, and they would not be charged for items or services that they did not purchase, which is the main goal of the project. The detection of fraudulent transactions will be accomplished using three machine-learning techniques: KNN, SVM, and Logistic Regression.
Final Exam Project Layout Introduction With more people using credit cards in their daily lives, credit card companies must take extra precautions to ensure the security and safety of their customers. The global credit card market revenue reached an impressive $152 billion in 2022, growing at a 9.1% annual rate, highlighting the massive scale of credit card transactions worldwide. The number of credit card holders worldwide has steadily increased to 1.25 billion in 2023, representing a 2.79% annual growth rate from 1.1 billion in 2018. (Caporal). Credit card holders in the United States number 166 million, while credit card holders in Canada number 36 million. Through the first half of 2023, 559,000 cases of identity theft were reported. With 440,666 reports, credit card fraud was the most common type of identity theft in 2022. In the first half of 2023, 219,713 credit card fraud reports were filed. The fastest-growing type of identity theft is synthetic fraud. There are two types of credit card fraud. The first is when an identity thief opens a credit card account in your name; reports of this fraudulent behavior increased 48% from 2019 to 2020. The second type is when an identity thief uses an existing account that you created, usually by stealing credit card information; reports on this type of fraud increased 9% from 2019 to 2020. These statistics piqued our interest because the numbers have been steadily increasing over the years, motivating us to try to solve the problem analytically by employing various machine learning methods to detect credit card fraudulent transactions within many transactions. (Radage) Project Goals The primary goal of this project is to detect credit card fraudulent transactions, as it is critical to identify fraudulent transactions so that customers are not charged for products they did not purchase. The detection of fraudulent credit card transactions will be performed using multiple ML techniques, and then a comparison will be made between the outcomes and results of each
Your preview ends here
Eager to read complete document? Join bartleby learn and gain access to the full version
  • Access to all documents
  • Unlimited textbook solutions
  • 24/7 expert homework help
Final Exam Project Layout technique to find the best and most suited model for the detection of fraudulent credit card transactions; graphs and numbers will also be provided. Furthermore, investigates previous literature and various techniques used to detect fraud within a dataset. Research Methodology CRISP-DM The CRISP-DM (Cross Industry Standard Process for Data Mining) is a process model that serves as the foundation for a data science process. It is divided into six stages: 1. Understanding the business - What does the business require? 2. Data comprehension - What data do we have/need? Is it sanitary? 3. Data preparation entails organizing the data to model it. 4. What modeling techniques should we employ? 5. Which model best meets the business objectives, according to the evaluation? 6. Deployment - How do stakeholders get their hands on the results? I believe that using CRISP-DM will make it easier to achieve efficient and elite results because it takes the project through the entire journey, beginning with understanding the business and data,
Final Exam Project Layout preparing the data, modeling it, and finally evaluating the model to ensure it is performing well. (Hotz) Phase 1: Business Awareness As previously stated, credit card fraud is increasing dramatically every year, and many people are facing the problem of having their credit cards breached by those fraudulent people, which is affecting their daily lives, as credit card payments are like taking out a loan. If the problem is not solved, many people will have large amounts of loans that they cannot repay, causing them to have a difficult life and not be able to afford necessary products; in the long run, not being able to repay the amount may result in them going to jail. Essentially, the proposed problem is the detection of fraudulent credit card transactions made by fraudsters to stop those breaches and prevent future breaches. Phase 2: Data Interpretation Because the model is based on a high-quality dataset, it is critical to obtain one during the Data understanding phase. The dataset will be explored by taking a closer look at the quality of the dataset, in addition to reading the description of the entire dataset and each attribute. It is also necessary to have a dataset with several mixed transaction types "Fraudulent and real" and a class to clarify the type of transaction, as well as identifiers to clarify the reason for the classification of 3 the transaction type. During the search for the best dataset, we will make sure to keep all those points in mind. Phase 3: Data Collection After selecting the most appropriate dataset, the preparation phase begins, which includes selecting the desired attributes or variables, cleaning it by excluding Null rows, deleting duplicated variables, treating outliers if necessary, and transforming data types to the desired
Final Exam Project Layout type. Data merging can also be performed, in which two or more attributes are merged. All these changes result in the desired outcome, which is that the data is now ready to be modeled. Phase 4: Modeling During the modeling phase, we will develop four machine-learning models KNN, SVM, Logistic Regression, and Nave Bayes. A comparison of the results will be presented later in the paper to determine which technique is best suited for detecting credit card fraudulent transactions. Phase 5: Evaluation and Implementation The final phase will present evaluations of the models by presenting their efficiency, accuracies, and any comments observed, to find the best and most suited model for detecting credit card fraud transactions.
Your preview ends here
Eager to read complete document? Join bartleby learn and gain access to the full version
  • Access to all documents
  • Unlimited textbook solutions
  • 24/7 expert homework help
Final Exam Project Layout Reference: 1. Caporal, Jack. “Identity Theft and Credit Card Fraud Statistics for 2023.” The Motley Fool , 6 Nov. 2023, www.fool.com/the-ascent/research/identity-theft-credit-card-fraud-statistics/ #:~:text=559%2C000%20instances%20of%20identity%20theft,growing%20form%20of %20identity%20theft. 2. Radage, Kalle. “The Growth of the Credit Card Industry in 2023.” Credit Card Processing and Merchant Account , 31 July 2023, www.clearlypayments.com/blog/growth-of-credit- cardindustryin2023/#:~:text=billion%20in%202018.,The%20number%20of%20credit %20card%20holders%20globally%20has%20steadily%20grown,from%201.1%20billion %20in%202018. 3. Hotz, Nick. “What Is CRISP-DM? - Data Science Process Alliance.” Data Science Process Alliance , 19 Jan. 2023, www.datascience-pm.com/crisp-dm-2.