A data-driven approach to predict the success of bank telemarketing

By Sérgio Moro, Paulo Cortez, Paulo Rita. 2014

999 citations recorded by the source corpus.

Abstract We propose a data mining (DM) approach to predict the success of telemarketing calls for selling bank long-term deposits. A Portuguese retail bank was addressed, with data collected from 2008 to 2013, thus including the effects of the recent financial crisis. We analyzed a large set of 150 features related with bank client, product and social-economic attributes. A semi-automatic feature selection was explored in the modeling phase, performed with the data prior to July 2012 and that allowed to select a reduced set of 22 features. We also compared four DM models: logistic regression, decision trees (DTs), neural network (NN) and support vector machine. Using two metrics, area of the receiver operating characteristic curve (AUC) and area of the LIFT cumulative curve (ALIFT), the four models were tested on an evaluation set, using the most recent data (after July 2012) and a rolling window scheme. The NN presented the best results (AUC = 0.8 and ALIFT = 0.7), allowing to reach 79% of the subscribers by selecting the half better classified clients. Also, two knowledge extraction methods, a sensitivity analysis and a DT, were applied to the NN model and revealed several key attributes (e.g., Euribor rate, direction of the call and bank agent experience). Such knowledge extraction confirmed the obtained model as credible and valuable for telemarketing campaign managers.

Subfields: Information Systems and Management, Marketing, Organizational Behavior and Human Resource Management, Statistics and Probability

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Information Systems and Management

In Marketing

In Organizational Behavior and Human Resource Management

In Statistics and Probability

Citations in the corpus

SplitNN-driven Vertical Partitioning (2020, methodology): “We evaluate our proposed method on three popular ﬁnancial datasets, namely Bank Marketing (Moro et al. , 2014a), Give me Some Credit Kaggle (2011a) and Financial PhraseBank (Malo et al. , 2014).”
Scalable Fair Clustering (2019, background): “The dataset5 is extracted from marketing campaigns of a Portuguese banking institution (Moro et al., 2014).”
An Interpretable Model with Globally Consistent Explanations for Credit Risk (2018, background): “decision trees and support vector machines for bank direct marketing [10, 9]), which cannot accommodate FICO’s monotonicity constraints.”
Learning Optimized Risk Scores (2016, background): “Dataset Source Classication Task adult Kohavi (1996) predict if a U.S. resident earns more than $50000 arrest Zeng et al. (2015) predict if a prisoner will be arrested within 3 years of release bank Moro et al. (2014) predict if person signs up for bank account after marketing call cancer Mangasaria”
Algorithmic Fairness (2020, methodology): “The Bank Marketing dataset is a publicly available dataset in the UCI repository [43, 103], and it includes 41,188 individuals with 20 attributes.”
Fair Clustering Through Fairlets (2018, background): “This dataset 3 contains one record for each phone call in a marketing campaign ran by a Portuguese banking institution (Moro et al. , 2014).”
Life-long learning based on dynamic combination model (2017, background): “The Mushroom dataset is comprised of the features of 23 species of mushroom, tagged with edible and poisonous [35].”
DART: Data Addition and Removal Trees (2020, neutral): “To test if R-DART works with other greedy split heuristics, we perform our deletion experiments on the Surgical, Vaccine, Adult, and Bank Marketing datasets using entropy as the splitting criterion.”
TabLLM: Few-shot Classification of Tabular Data with Large Language Models (2022, neutral): “We analyzed the following datasets: • Bank (Kadra et al., 2021) contains information of a direct marketing campaign from a Portugese banking institution (Moro et al., 2014).”
Does mitigating ML's impact disparity require treatment disparity? (2017, background): “Income UCI [22] Gender (female) income > $50k 32,561 Marketing UCI [23] Status (married) customer subscribes 45,211 Credit UCI [24] Gender (female) credit card default 30,000 Employee Attr.”
Machine Unlearning for Random Forests (2020, background): “• Bank Marketing (Moro et al., 2014; Dua & Graff, 2019) consists of 41,188 marketing phone calls (4,640 positive) from a Portuguese banking institution.”
The Eﬀects of Data Quality on Machine Learning Performance (2022, background): “It contains different characteristics of a person and has a binary target stating whether a term deposit is subscribed [45, 55].”
Measuring Bias (2023, neutral): “The protected group is marital status [25].”
SAINT: Improved Neural Networks for Tabular Data via Row Attention and Contrastive Pre-Training (2021, neutral): “The Bank Marketing dataset is also compiled and organized by UCI and released for research use [29].”
A Review on Fairness in Machine Learning (2022, methodology): “[59, 152, 184] Marketing 41,188 Age Whether the client subscribed to a”
Predicting the success of bank telemarketing using deep convolutional neural network (2015, background): “Data are the benchmark data collected and donated publicly by Moro [8].”
Counterfactual Explanations and Algorithmic Recourses for Machine Learning: A Review (2020, neutral): “• Tabular: Adult income, German credit, Student Performance, Breast cancer, Default of credit, Shopping, Iris, Wine, Spambee, Covertype, ICU [87], LendingClub [294], Give Me Some Credit [162], COMPAS [155], LSAT [36], Pima diabetes [283], HELOC/FICO [100], Fannie Mae [208], Portuguese Bank [223], Sa”