A SMOTe based Oversampling Data-Point Approach to Solving the Credit Card Data Imbalance Problem in Financial Fraud Detection

Mqadi, Nhlakanipho Michael; Naicker, Nalindren; Adeliyi, Timothy T.

doi:10.12785/ijcds/100128

Cited by 16 publications

(8 citation statements)

References 24 publications

(28 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The efficacy of this approach was evaluated using a Principal Component Analysis (PCA)--based Artificial Neural Network (ANN) classifier, demonstrating its superiority over existing methods. On a parallel note, Mqadi et al [11] introduced a data-point machine-learning technique to address the challenges posed by an imbalanced credit card database.…”

Section: IImentioning

confidence: 99%

Feature Selection Techniques for Enhancing Credit Card Fraud Detection Performance: A Hybrid Metaheuristic Approach Using Nature-Inspired Algorithms

Sure,

Pandey,

Parnami

et al. 2024

IJRASET

View full text Add to dashboard Cite

The surge in credit card usage has led to a parallel increase in fraudulent transactions, necessitating advanced detection systems. Traditional methods encounter challenges with high-dimensional and imbalanced datasets. This paper proposes a comprehensive approach integrating metaheuristic algorithms and deep learning techniques to enhance fraud detection accuracy. The first contribution, the Rock Hyrax Swarm Optimization Feature Selection (RHSOFS) algorithm, draws inspiration from the collective behaviors of rock hyrax swarms to effectively select relevant features from high-dimensional datasets. RHSOFS identifies an optimal subset of features critical for fraud detection through supervised machine learning. Complementing this, the second contribution leverages a hybrid deep learning model. Beginning by organizing transactional data and constructing a Logical Graph of Behavior Profile (LGBP) to abstract transaction details, the Modified Butterfly Optimization Algorithm (MBOA) selects important features from the dataset. The hybrid model, integrating Convolutional Neural Networks (CNN) and Recurrent Neural Networks (RNN), establishes rational connections between transactional characteristics, enhancing detection performance. To evaluate the approach, comparative efficiency analyses against existing methods, including Differential Evolutionary Feature Selection (DEFS), Genetic Algorithm Feature Selection (GAFS), Particle Swarm Optimization Feature Selection (PSOFS), and Ant Colony Optimization Feature Selection (ACOFS), are conducted. Results demonstrate the superiority of the approach in terms of both reliability and recognition rate, validated through rigorous statistical testing. The proposed hybrid approach signifies a significant advancement in credit card fraud detection systems, offering enhanced accuracy and efficiency in combating fraudulent activities amidst the growing complexity of transactional datasets.

show abstract

Section: IImentioning

confidence: 99%

Feature Selection Techniques for Enhancing Credit Card Fraud Detection Performance: A Hybrid Metaheuristic Approach Using Nature-Inspired Algorithms

Sure,

Pandey,

Parnami

et al. 2024

IJRASET

View full text Add to dashboard Cite

show abstract

“…Decision Tree The decision tree [38] is a commonly used machine learning algorithm that uses a rule-based tree structure to split data into predefined classes. The decision rules for splitting the data are based on the characteristics and classification of the data set.…”

Section: Machine Learning Algorithmsmentioning

confidence: 99%

Analysis of Road Traffic Accidents Severity Using a Pruned Tree-Based Model

Adeliyi,

Oluwadele,

Igwe

et al. 2023

IJTDI

View full text Add to dashboard Cite

Traffic accidents are becoming a global issue, causing enormous losses in both human and financial resources. According to a World Health Organization assessment, the severity of road accidents affects between 20 and 50 million people each year. This study intends to examine significant factors that contribute to road traffic accident severity. Seven machine learning models namely, Naive Bayes, KNN, Logistic model tree, Decision Tree, Random Tree, and Logistic Regression machine learning models were compared to the J48 pruned tree model to analyze and predict accident severity in the road traffic accident. To compare the effectiveness of the machine learning models, ten well-known performance evaluation metrics were employed. According to the experimental results, the J48 pruned tree model performed more accurately than the other seven machine learning models. According to the analysis, the number of casualties, the number of vehicles involved in the accident, the weather conditions, and the lighting conditions of the road, is the main determinant of road traffic accident severity.

show abstract

“…Additionally, for solving the multiclassification problem of financial price movement prediction, the training samples of each class are usually unbalanced, which leads to biased classification results and low accuracy [30,31]. The Synthetic Minority Oversampling Technique (SMOTE), which was proposed by Chawla et al [32], is an efficient method for solving unbalanced samples by oversampling the minority [33], and it has been successfully and widely applied in many fields [33][34][35][36]. Therefore, following the research of Chawla et al [32], the SMOTE-based approach is employed and integrated into the proposed method to balance the model training samples of different classes before the model training of fuzzy rough set (FRS).…”

Section: Related Workmentioning

confidence: 99%

Financial Futures Prediction Using Fuzzy Rough Set and Synthetic Minority Oversampling Technique

Zhu

Liu

2022

Advances in Mathematical Physics

View full text Add to dashboard Cite

In this research, a novel approach called SMOTE-FRS is proposed for movement prediction and trading simulation of the Chinese Stock Index 300 (CSI300) futures, which is the most crucial financial futures in the Chinese A-share market. First, the SMOTE- (Synthetic Minority Oversampling Technique-) based method is employed to address the sample unbalance problem by oversampling the minority class and undersampling the majority class of the futures price change. Then, the FRS- (fuzzy rough set-) based method, as an efficient tool for analyzing complex and nonlinear information with high noise and uncertainty of financial time series, is adopted for the price change multiclassification of the CSI300 futures. Next, based on the multiclassification results of the futures price movement, a trading strategy is developed to execute a one-year simulated trading for an out-of-sample test of the trained model. From the experimental results, it is found that the proposed method averagely yielded an accumulated return of 6.36%, a F1-measure of 65.94%, and a hit ratio of 62.39% in the four testing periods, indicating that the proposed method is more accurate and more profitable than the benchmarks. Therefore, the proposed method could be applied by the market participants as an alternative prediction and trading system to forecast and trade in the Chinese financial futures market.

show abstract

A SMOTe based Oversampling Data-Point Approach to Solving the Credit Card Data Imbalance Problem in Financial Fraud Detection

Cited by 16 publications

References 24 publications

Feature Selection Techniques for Enhancing Credit Card Fraud Detection Performance: A Hybrid Metaheuristic Approach Using Nature-Inspired Algorithms

Feature Selection Techniques for Enhancing Credit Card Fraud Detection Performance: A Hybrid Metaheuristic Approach Using Nature-Inspired Algorithms

Analysis of Road Traffic Accidents Severity Using a Pruned Tree-Based Model

Financial Futures Prediction Using Fuzzy Rough Set and Synthetic Minority Oversampling Technique

Contact Info

Product

Resources

About