The problem for this course project is related to direct marketing campaigns (phone calls) of a Portuguese banking institution. The classification goal is to predict whether the client subscribes a term deposit or not. The target class is the last attribute (subscribed) and has two values (yes and no).

The training set is in trainset.csv and the test set is in testset.csv. The dataset contains subscribed (yes) and unsubscribed (no) customers.

Attribute Information:

1. age (numeric)

2. job: type of job (categorical: 'admin.', 'blue-collar', 'entrepreneur', 'housemaid', 'management',

'retired', 'self-employed', 'services', 'student', 'technician', 'unemployed', 'unknown')

3. marital: marital status (categorical: 'divorced', 'married', 'single', 'unknown'; note: 'divorced'

means divorced or widowed)

4. education: (categorical: 'basic.4y', 'basic.6y', 'basic.9y', 'high.school', 'illiterate',

'professional.course', 'university.degree', 'unknown')

5. housing: has housing loan? (categorical: 'no', 'yes', 'unknown')

6. loan: has personal loan? (categorical: 'no', 'yes', 'unknown')

7. contact: contact communication type (categorical: 'cellular', 'telephone')

8. month: last contact month of year (categorical: 'jan', 'feb', 'mar', ..., 'nov', 'dec')

9. day_of_week: last contact day of the week (categorical: 'mon', 'tue', 'wed', 'thu', 'fri')

10. duration: last contact duration, in seconds (numeric).

11. campaign: number of contacts performed during this campaign and for this client (numeric,

includes last contact)

12. pdays: number of days that passed by after the client was last contacted from a previous

campaign (numeric; 999 means client was not previously contacted)

13. poutcome: outcome of the previous marketing campaign (categorical: 'failure', 'nonexistent',

'success')

14. nr.employed: number of employees - quarterly indicator (numeric)

15. Target Attribute: Subscribed - has the client subscribed a term deposit? (binary: 'yes','no')

Steps

The project involves the following steps:

1. Data exploration: try to know data and represents statistics for the important features among the features and the target attribute.

2. Preprocessing the data. The goal of this step is to extract features from records in the training set and use these features to test data sets. Note that the data have unknown values need to be

cleaned.

3. Use classification-learning methods provided by Python to learn models from the set of training examples. You can use any of the classification methods (decision tree, regression, KNN, ) for this purpose. You should build at least two different and fundamentally distinct models (e.g., one decision tree and one regression). Note that two decision trees with different attributes count as one model.

4. Test the learned models on the test set and report the testing results.

Important note: For the final course project, I expect you to be able to find techniques that may not have been directly covered in the labs. This is because by now you have a decent understanding of Python and the labs have prepared you to find answers from the Web. For example, you should be able to look at some codes related to data understanding, visualization, and attribute and feature selection.

Here are some keywords you can search on Google and check the first few pages:

data exploration python

data understanding python

feature selection python