Lending Club Data Science Demo

This internal DXC demonstration aims at presenting a full Data Science project on a peer-to-peer lending problematic. Starting with the raw data collection and ending with a minimal viable application, it presents the typical deliverables that will be submitted to the client in the Predict offer.

Access project and data

git clone https://gitlab.com/dxc_data_science/Lending_Club.git

Cloning the repo, you will get everything you need to explore the project except the most important thing: the Data (The file is too large to be stored on GitLab).

To access the data you will have to:

1. Create a folder called inputs in the Lending_Club folder
2. Download the loan.csv file from Kaggle and put it in the inputs folder

DataBook (Exploratory Data Analysis)

A Data Science project always starts exploring the data in order to visualize, clean and understand it. It is also the best way to validate some assumptions and be prepared for the modelling stage.

Our project includes 2 files that presents the data and the exploration:

• LCDataDescription_final.xlsx located in Lending_Club\data_description describes all the variables in the dataset (Description sheet) and presents descriptive statistics on quantitative variables (Summary statistics sheet).
• LendingClubRmd.Rmd located in Lending_Club\eda\LendingClubRmd is the source file that generates the DataBook in html format.

prerequisites for generating html report

• R - click here if not installed
• R studio - click here if not installed
• R dxcrmd package - click here if not installed

In order to generate the html report: Open LendingClubRmd.Rmd in RStudio and click on Knit.

Enjoy the reading!

Machine Learning Model

InsightBook

Application