Skip to content

Latest commit

 

History

History
117 lines (85 loc) · 5.88 KB

README.md

File metadata and controls

117 lines (85 loc) · 5.88 KB
Raw

Introduction

Company Introduction - AE Corp

Our client for this project is a retail banking institution.

  • They are going to float a stock trading facility for their existing customers.
  • The idea is to use data to classify whether a customer belongs to a high net worth or low net worth group.
  • They will have to incentivize their customers to adopt their offerings.
  • One way to incentivize could be to offer discounts on the commission for trading transactions.

Current Scenario

The company rolled out this service to about 10,000+ of its customers and observed their trading behavior for 6 months and after that, they labeled them into two revenue grids 1 and 2.

Problem Statement

The current process suffers from the following problems:

  • Only about 10% of the customers do enough trades for earnings after discounts to be profitable.
  • The company wants to figure out, which are those 10% customers so that it can selectively offer them a discount.

Project Details

  • Co-ordinating with - The marketing department to supplement their campaigns with a more proactive approach.
  • Our Role - To build a classification model using the datasets.
  • Project Deliverables - Predict whether a customer belongs to a high net worth or low net worth group.
  • Machine Learning Task: Classification
  • Target Variable: Status (high net worth (1) / low net worth (2))
  • Win Condition - No machine learning model in the company for this currently, therefore no quantifiable win condition. We just need to build the best possible model.
  • Evaluation Metric - The model evaluation will be based on the F1 Score score.

Data Acquisition & Description

We are provided with a dataset containing all the necessary information about the customers like their occupation, family income, gender, region, balance transfer, children, etc.

Also included in the dataset is the column Revenue_Grid which classifies the customers into high net worth customers (1) and low net worth customers (2).

This is the data that we have to predict for future customers.

  • Train Set: The train set contains 8124 rows and 32 columns.The last column Revenue_Grid is the target variable.

  • Test Set: The test set contains 2031 rows and 31 columns.The test set doesn’t contain the Revenue_Grid column which need to be predicted

  • The Dataset contains the following columns:

Id Feature Description
01 REF_NO Reference Number of the customer.
02 children Number of children each customer has.
03 Age_band Age Group to which the customer belongs.
04 status Marital Status of the customer.
05 occupation Job or profession of the customer.
06 occupation_partner Job or profession of the customer's partner.
07 home_status Home Status of the customers.
08 family_income Income Range of the customer's family.
09 self_employed Whether self-employed or not.
10 self_employed_partner Whether the partner self-employed or not.
11 year_last_moved Moving Year from the last location of the customer.
12 TVarea Television Region of the customer.
13 post_code Postal Code of the customer.
14 post_area Postal Area of the customer.
15 Average_Credit_Card_Transaction Average Credit Card Transaction per year by the customer.
16 Balance_Transfer Transfer of the Balance in an account to another account by the customer.
17 Term_Deposit Cash Investment Help at Financial Institute provided to the customer.
18 Life_Insurance Basic Life Insurance Coverage of the customer.
19 Medical_Insurance Medical Insurance Coverage of the customer.
20 Average_A/C_Balance Average Balance in the account of the customer.
21 Personal_Loan Amount of Personal Loan taken by the customer.
22 Investment_in_Mutual_Fund Amount Invested in Mutual Funds by the customer.
23 Investment_Tax_Saving_Bond Amount Invested in Tax Saving Bond by the customer.
24 Home_Loan Amount of Home Loan taken by the customer.
25 Online_Purchase_Amount Amount spent by the customer on online purchases.
26 gender Gender of the customer.
27 region Religion of the customer.
28 Investment_in_Commudity Amount Invested in Commodity by the customer.
29 Investment_in_Equity Amount Invested in Equity by the customer.
30 Investment_in_Derivative Amount Invested in Derivatives by the customer.
31 Portfolio_Balance Balanced Investment Strategy of the customer.
32 Revenue_Grid Grid report of the customers.

Data Profiling & Pre-Processing

Model Summarization

Submission

The Client requested the Submission file with below mentioned conditions -

  • The submission file should be in the csv format.
  • It should have 2031 rows.
  • It should only have 2 columns, 1st column: REF_NO values, and 2nd column: predicted Revenue_Grid values.
  • The submission file should not have column names in the first row.

Based on the requirments, the file was generated whichcan be viewed here

Thank you