Dark Mode
Client Subscription Model Data
Finance & Banking Analytics
Tags and Keywords
Trusted By
"No reviews yet"
Free
About
Data detailing direct marketing campaigns conducted by a Portuguese banking institution, primarily via phone calls. The primary objective is predicting the success of these telemarketing efforts, specifically whether a client will subscribe to a bank term deposit product. The records capture client demographics, attributes related to the last contact made during the current campaign, and relevant social and economic context indicators.
Columns
The most detailed version of this data contains 21 variables, including the target output:
- age: The client's age (numeric).
- job: Type of job (e.g., 'admin.', 'blue-collar', 'management').
- marital: Marital status ('divorced', 'married', 'single'). Note that 'divorced' includes widowed status.
- education: Client's educational level (e.g., 'high.school', 'university.degree').
- default, housing, loan: Binary/categorical flags indicating if the client has credit in default, a housing loan, or a personal loan, respectively.
- contact: Communication type used for the last contact ('cellular' or 'telephone').
- month, day_of_week: The time of the last contact (e.g., 'jan' or 'mon').
- duration: The duration of the last contact in seconds (numeric). Note: This variable highly correlates with the output and should be excluded for building realistic predictive models, as its value is unknown before the call ends.
- campaign: The total number of contacts performed during the current campaign for this client.
- pdays: The number of days passed since the client was last contacted from a previous campaign (999 indicates no prior contact).
- previous: The number of contacts performed before this campaign.
- poutcome: The outcome of the previous marketing campaign ('failure', 'success', 'nonexistent').
- emp.var.rate, cons.price.idx, cons.conf.idx, euribor3m, nr.employed: Quarterly, monthly, or daily indicators reflecting the social and economic context (e.g., employment variation rate, consumer price index, euribor 3 month rate).
- y: The output target variable: whether the client subscribed to a term deposit ('yes' or 'no').
Distribution
Four datasets are available, offering different sizes and input features. The largest and most detailed file,
bank-additional-full.csv, contains 41,188 examples and 20 input variables, spanning a date range from May 2008 to November 2010. Smaller subsets, such as bank-additional.csv (10% of the full sample, 4,119 examples), are included to facilitate testing of computationally demanding machine learning algorithms like SVM. Older versions of the dataset with 17 inputs are also provided. Specific detailed metrics (like means, standard deviations, or valid counts) for all variables are currently not available in the supporting documentation.Usage
Ideal applications for this data include:
- Developing machine learning models to classify potential term deposit subscribers.
- Benchmarking various prediction algorithms (e.g., using the 'duration' attribute for benchmark testing only).
- Analyzing the effectiveness of direct marketing strategies based on client attributes and campaign history.
- Exploring the impact of social and economic context indicators on financial product subscription rates.
Coverage
The data originates from a Portuguese banking institution and covers interactions and client attributes recorded between May 2008 and November 2010. The social and economic variables provide context spanning this period.
License
CC0: Public Domain
Who Can Use It
- Data Scientists and ML Researchers: Building and comparing classification models for binary prediction tasks.
- Financial Analysts: Studying factors that drive customer conversion rates for time deposits.
- Marketing Strategists: Optimising future telemarketing campaign structures and targeting based on historical success metrics.
Dataset Name Suggestions
- Bank Term Deposit Prediction
- Portuguese Bank Telemarketing Success
- Client Subscription Model Data
- Direct Marketing Outcome Data
Attributes
Original Data Source:Client Subscription Model Data
QUALITYLoading...