Jul 2026· An International Journal of Optimization and Control: Theories & Applications (IJOCTA)· Vol 16, pp. 1605· 0 citations
TL;DR
Two overarching results are revealed: near-perfect performance under a full feature set is largely attributable to target leakage rather than a genuine predictive signal, and, under deployment, valid early-warning features degrade cost-sensitive performance relative to no resampling when SMOTE is used.
Abstract
The Synthetic Minority Oversampling Technique (SMOTE) has become the default preprocessing step for handling class imbalance in software effort-risk prediction, yet its effectiveness in this domain has not been rigorously tested. This study addresses that gap through a controlled factorial experiment on the publicly available Software Development Effort Dataset Annotated with Expert Estimates dataset, comprising 4,329 software issues from Apache projects drawn from an initial repository of 23,186 records. Six resampling strategies are compared across four classifier families under both standard and cost-sensitive evaluation metrics that weight missed high-risk issues more heavily than false alarms. A secondary contribution is the analysis of how resampling interacts with the extreme class imbalance characteristic of real-world effort data (2.2% minority rate), a regime substantially more severe than those examined in prior investigations. Results are interpreted through SHAP-based feature attribution to determine whether oversampling alters which features the models rely on. The findings reveal two overarching results: near-perfect performance under a full feature set is largely attributable to target leakage rather than a genuine predictive signal, and, under deployment, valid early-warning features degrade cost-sensitive performance relative to no resampling when SMOTE is used. Cost-sensitive weighting emerges as the more reliable alternative, preserving both performance and feature attribution structure. These findings challenge the uncritical adoption of SMOTE in software analytics and carry direct implications for the design of reproducible, interpretable risk-detection pipelines.
RiskBlend is proposed, a classifier-agnostic prioritization framework that combines four complementary risk signals: historical failure patterns, prediction shift, decision-boundary shift, and neighborhood change that achieves the highest average APFD in all 80 dataset-classifier-scenario combinations.
Accurate credit risk prediction plays a critical role in strengthening the financial stability of microfinance institutions, especially in developing economies, where increasing loan defaults and imbalanced borrower records create significant challenges for reliable decision-making. Although machine learning approaches have improved credit assessment practices, existing models often favor majority-class borrowers and fail to detect high-risk default cases effectively because of severe class imbalance. This limitation highlights the need for more robust and imbalanced-sensitive predictive frameworks. This study aims to develop an effective machine learning-based credit risk prediction framework by integrating data balancing strategies with ensemble and deep learning models. This study systematically investigates the impact of baseline learning and multiple resampling techniques, including oversampling, undersampling, and hybrid methods, when applied to Random Forest, XGBoost, LightGBM, CatBoost, and Deep Neural Network classifiers. The effectiveness of the proposed models was assessed using imbalance-aware evaluation measures, particularly ROC-AUC and Geometric Mean, along with conventional classification metrics. The experimental findings demonstrate that incorporating resampling techniques substantially improved the default detection performance. The DNN model combined with SMOTETomek achieved the best results, obtaining 94.9% F1-score, ROC-AUC 98%, and 97.2% of G-Mean. CatBoost also exhibited consistent competitiveness across different sampling configurations. These findings suggest that hybrid sampling integrated with advanced learning architectures can provide a reliable and practical solution for managing credit risk in imbalanced microfinance datasets, supporting improved lending decisions and sustainable financial operations in the future.
Class-imbalance handling is routinely evaluated on a single benchmark dataset, and the resulting conclusions are reported as if they were properties of the method. We show this practice is unsafe. On the public Kaggle credit-card fraud dataset, under a leakage-free nested cross-validation protocol in which the decision threshold is selected on a held-out inner validation fold, a plain Random Forest at the default 0.5 threshold attains F1 = 0.861 +/- 0.021, and threshold tuning yields it no benefit (delta-F1 = -0.002). Read alone, this supports an appealing conclusion: for a well-calibrated ensemble, imbalance handling is unnecessary. We then apply the identical protocol to 45 binary tasks spanning imbalance ratios from 1:1.5 to 1:178 (2,025 model fits, four model families). The conclusion reverses. Random Forest benefits most from threshold tuning across the suite (delta-F1 = +0.101 +/- 0.134), not least, while three other families replicate their fraud-dataset behaviour almost exactly. SMOTE likewise harms the fraud dataset but helps across the suite (mean delta-F1 = +0.076; 138 wins, 39 losses; Wilcoxon p = 2.7e-17). Two further results. Threshold-tuning benefit is non-monotonic in the imbalance ratio: near zero below 1:5, peaking at +0.120 in the 1:15-1:40 band, declining to +0.045 beyond 1:100 - explaining why the fraud dataset, at 1:577, is an unrepresentative place to study the question. And we reject an intuitive heuristic: validation-set calibration error does not predict tuning benefit (expected calibration error r = -0.087; Brier r = +0.137), so calibration diagnostics cannot tell a practitioner whether tuning is worthwhile. We release the protocol, the 45-task harness, and all per-run metrics.
Credit card fraud is a menace to financial institutions, but detection is compromised by highly imbalanced transaction datasets. This study proposes an advanced machine learning framework optimized for fraud detection. To address the issue of data imbalance, SMOTE-Tomek Links is applied to synthetically generate minority fraud cases while removing noisy, overlapping majority-class instances. Recursive Feature Elimination (RFE) is deployed to identify the optimal features, and RandomizedSearchCV automates hyperparameter optimization. The study introduces a Stacked Logistic Regression ensemble to combine the predictive capacity of optimized Random Forest and XGBoost base classifiers. The model’s effectiveness is assessed using seven evaluation methods: accuracy, recall, precision, confusion matrix, F1-score, Receiver Operating Characteristic Area Under the Curve (ROC-AUC) score and the Area Under the Precision-Recall Curve (AUC-PR) score. Findings reveal that the proposed stacked model performance surpasses both individual base models. While achieving deceptively high baseline accuracy across all models, the stacked ensemble delivers a superior AUC-PR score of 0.8207 and an F1-score of 0.93. This minimizes the confusion matrix misclassifications to just 20 False Negatives and 5 False Positives. The framework provides a cost-optimized operational engine that aggressively mitigates bank fraud losses while successfully shielding legitimate cardholders from accidental checkout declines.
Uduh Israel Akakoh, G. N. Edegbe· FUDMA Journal of Sciences· 0 citations
Despite the growing dominance of deep neural architectures, logistic regression continues to hold its ground as a go-to
tool for binary classification — largely because of how easy it is to interpret and how cheaply it runs. That said, it is far from
bulletproof. Certain data conditions can quietly erode its performance in serious ways. This paper takes a hands-on, experimentdriven approach to investigating four well-known but often under-examined failure scenarios: multicollinearity among input
features, decision boundaries that curve rather than cut straight, the convergence breakdown that comes with complete
separation in high-dimensional spaces, and the silent damage caused by heavily lopsided class distributions. We ran experiments
on purpose-built synthetic datasets as well as two widely used real-world benchmarks from the UCI repository — the Wisconsin
Breast Cancer diagnostic set and the Credit Card Fraud Detection collection — measuring outcomes through ROC-AUC and
F1-score. On top of documenting where things fall apart, we also tested a suite of targeted fixes: L1 and L2 penalization,
polynomial feature expansion, and the SMOTE oversampling method. The numbers paint a clear picture. Without any
intervention, logistic regression barely beats a coin flip on non-linear data (ROC-AUC of 0.52) and essentially ignores the
minority class under extreme skew (F1 below 0.10). But each of the tested remedies brought meaningful recovery — polynomial
expansion pushed non-linear classification up to 0.97 ROC-AUC, while SMOTE lifted minority-class F1 to 0.82. We distill these
findings into a straightforward decision guide that practitioners can use to quickly diagnose what is going wrong with their
logistic regression model and choose the right corrective action.
Afaan Hussain Shah, Saqib Anjum Masoodi· International Journal for Re...· 0 citations
The problem of class imbalance in machine learning classification is widely present and difficult across the machine learning area, especially in real-world tasks, including fraud detection, medical diagnosing, network intrusion detection and fault prediction. When this occurs, the minority population is more likely to capture the important occurrences and the conventional machine learning models normally focus on the majority population and give misleading accuracy with poor generalization and high costs of misclassification. This paper is the result of an extensive research into powerful machine learning techniques in the classification of imbalanced datasets. The paper presents a systematic review of theoretical underpinnings of learning imbalance, literature reviews on state-of-the-art methods, such as data, algorithm-level and ensemble based methods, and suggests a convergent system methodology to build a robust classifier. Linear resampling algorithms, cost-effective learning algorithms, hybrid ensemble algorithms, and imbalanced data evaluation metrics are discussed in details. An organized experimental procedure is described to measure robustness when imbalance ratios and various noise levels are changing. Comparative findings indicate that hybrid methods that combine adaptive resampling and cost sensitive loss functions are always better than simpler classifiers based on their F1-score, G-mean, and area under the precision-recall curve. The discussion demonstrates practical trade-offs between model performance, model complexity and interpretability. In the conclusion part, the paper highlights future research directions which include scalable imbalance learning, deep learning adaptations and domain aware evaluation strategies. The paper is an excellent source of information to a researcher and practitioner aiming at finding principled and effective solutions to imbalanced classification problems.
Vasanth Kumar, Girija Rajan· International Journal of Mac...· 0 citations
