1    Keynotes

• Grading: 50% of your final grade.

• Deadline: 23:59 on 16th of April, 2023.

• Submission files:  1) a .m file which consists of your coding; 2) a .mat file which consists of your estimated results. 3) a .pdf file which consists of your detailed report.

2    Data and Objectives

LendingClub is an American peer-to-peer lending company, currently the world’s largest platform that allows for individuals to both invest and borrow on the platform. Borrowers can obtain unsecured personal loans from the platform, and this coursework is set up for you to assess your ability to predict defaulters in the data using the predictors provided in the data. The data is a random sample of loans issued on the platform between 2007 and 2015, including the loan status (the response variable) and payment information (predictors).

You have been provided with three .csv data files:  trainData:  This is the dataset on which you will train all your models.   testData:  This is the dataset on which you will evaluate your models’ performance.   variable Description:  This file provides more information regarding each variable.

3    Coursework instructions

You are also provided with a .m file.  Some of the coding, for instance, importing and cleaning data are prepared for you already in this file.  You are required to fill the codes between lines of ######.  What you have learned in the class should be sufficient for you to complete this task. To maximise your chance to get high scores I would advise you to follow these rules:

1. comment your code using ’%’and keep your comment concise;

2. use consistent indent style for your coding. Codes in the same block should have the same indent;

3. save figures as .png file, and use them in your PDF report.

4. save all required variables in one .mat file.

5. your PDF report shouldn’t exceed 4 pages, excluding figures and appendix. So, it is crucial to keep your sentences and paragraphs short and informative.

6. copy your coding script (i.e. the content in .m file) and attach it at the end of your reporting PDF file as the appendix.

7. you can use internet source of materials (although not necessarily encouraged), but it has to be digested and should be consistent with the subjects we covered in lectures. If you use external materials that suggest you don’t understand the concepts you are talking about, or if I find that the materials are directly copy-and-pasted, your grading will be penalized.

8. save .m .mat .pdf files in the following format:  for example a student called John Smith with student ID of 123456, should save his files as John Smith 123456.m, John Smith 123456.mat and John Smith 123456.pdf.

9. in your PDF report, set the font size of the main body text, as well as the appendix, as 12.

10. some suggestions on the structure of your PDF report - it can consist of the following sections:

(a) Section One: Introduction

give a very brief introduction on the problem you are investigating, and the models you will consider to tackle this problem. Give a general “big picture” of your findings and your conclusion. (should be around a page or less);

(b) Section Two: Methodologies

give detailed accounts of methods you used to import data, clean data, as well as models you employed to analyse the dataset.  briefly explain your models and the underlying theories. Point out pros and cons of each model, as well as any issues you have encountered with various models, and how you solved those problems.

(c) Section Three: Main findings

in this section you should present your estimation results for each model and compare between them and draw your conclusion. Include all your figures and tables you may have obtained here.