Enhancing the Random Forest Model via Synthetic Minority Oversampling Technique for Credit-Card Fraud Detection

Abstract

Fraudsters increasingly exploit unauthorized credit card information for financial gain, targeting un-suspecting users, especially as financial institutions expand their services to semi-urban and rural areas. This, in turn, has continued to ripple across society, causing huge financial losses and lowering user trust implications for all cardholders. Thus, banks cum financial institutions are today poised to implement fraud detection schemes. Five algorithms were trained with and without the application of the Synthetic Minority Over-sampling Technique (SMOTE) to assess their performance. These algorithms included Random Forest (RF), K-Nearest Neighbors (KNN), Naïve Bayes (NB), Support Vector Machines (SVM), and Logistic Regression (LR). The methodology was implemented and tested through an API using Flask and Streamlit in Python. Before applying SMOTE, the RF classifier outperformed the others with an accuracy of 0.9802, while the accuracies for LR, KNN, NB, and SVM were 0.9219, 0.9435, 0.9508, and 0.9008, respectively. Conversely, after the application of SMOTE, RF achieved a prediction accuracy of 0.9919, whereas LR, KNN, NB, and SVM attained accuracies of 0.9805, 0.9210, 0.9125, and 0.8145, respectively. These results highlight the effectiveness of combining RF with SMOTE to enhance prediction accuracy in credit card fraud detection.

StreamBoostE: A Hybrid Boosting-Collaborative Filter Scheme for Adaptive User-Item Recommender for Streaming Services

“…The ensemble has benefits [136]: (a) it yields more features via user-based and item-based collaborative filtering approach that ensures faster hybrid model construction and training [137], (b) the adapted collaborative filtering approach ensured lessened training time for the XGBoost ensemble especially its fusion with streamlit and flask to yield further integration over web-contents, where quick is critical [138], [139], (c) it yielded faster implementation with robust and effective crosschannel apps/platforms integration [140], (d) XGBoost yields enhanced accuracy in that adapted feats did not degrade performance compared to [46], [135]. Our ensemble successfully predicted and recommended targeted-user items for streaming media contents transactions for the Netflix multimedia streaming platforms [141], [142] with minimal false-positives.…”

“…Machine learning (ML) models learns these patterns via features of interest, which helps them identify these patterns as signature classification that deviates from the norm [40]. A variety of ML have yielded resultant success with its adoption in collaborative filtering algorithm to include: Logistic Regression [41]- [43], Deep Learning [44]- [46], Bayesian model [47]- [49], Support Vector Machine [50]- [52], Random Forest [53]- [55], K-Nearest Neighbors [56]- [58], and in other models [59]- [61]. Their flexibility and performance is greatly hampered/degraded with the adopted choice in feature selection technique and data-preprocessing scheme [62], [63].…”

CoDuBoTeSS: A Pilot Study to Eradicate Counterfeit Drugs via a Blockchain Tracer Support System on the Nigerian Frontier

“…The ensemble divides the roles into five (5) as represented via the 5chaincodes on the distributed hyper fabric ledger [135] technology to effectively handle the business transaction logic on the blockchain [136]. The model control was deployed via chaincode permissions and encryption mechanisms to enhance data security and privacy control for the support system traceability model [137].…”

Counterfeit Drugs Detection in the Nigeria Pharma-Chain via Enhanced Blockchain-based Mobile Authentication Service

“…The ensemble divides stakeholder roles into 5 represented via 5-chaincodes on the distributed hyper-ledger fabric technology to effectively handle the business transaction logic on the blockchain. The model control was deployed via chaincode permission and encryption mechanism to enhance data security and privacy control for the support system traceability model [86]. The resulting model showed a low response time to the query request, alongside stable time convergence for the system throughput as supported by [137].…”

“…With over 2480 healthcare facilities assessed in all the 6-Geo zones in Nigeria, Oyetunde [38] noted that availability of germane/basic drugs were found to be as low as about 25.2percent. Thus, stock-outs of basic medicine in primary healthcare facilities in Nigeria is almost a norm [86]. 5.…”