Bank Marketing Customer Conversion Data for Imbalanced Classification

13.5 KB of data in XLS format accompanies a study on forecasting customer conversion in bank marketing. The dataset, authored by Fahim Nasir and last updated in May 2026, is used to evaluate models and sampling techniques under a proposed quadrilateral evaluation framework. It supports research into responsible predictive analytics for Banking 4.0.

Use Cases

• Evaluating model-sampling combinations for imbalanced classification based on the described quadrilateral framework.
• Benchmarking ensemble and deep learning models on bank marketing data as mentioned in the study.
• Assessing the trade-offs of synthetic data generation techniques like Borderline2SMOTE on model performance.
• Identifying key drivers of customer responses using SHAP explanations for predictive inference.
• Developing transparent and reproducible model selection strategies for responsible banking analytics.

Strengths

• Dataset is 13.5 KB in size, indicating a focused and manageable scope for analysis.
• Accompanies a detailed methodological study proposing a two-stage evaluation framework for imbalanced classification.

Limitations

• Row count is unknown, which may limit suitability assessment for large-scale modeling.
• Column-level documentation is absent; field semantics must be inferred after download.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

figshare

Collection Method

Likely collected or generated as part of a research study on bank marketing and imbalanced classification.

Freshness

Last updated 2026-05-15 17:43:46; freshness should be verified.