Overview

Applied Machine-Learning and Big-Data for Credit Risk

About this course

This course aims to provide a comprehensive framework to understand modern credit applications. To achieve this goal, students must get familiarized with state-of-the-art technologies in the field of data engineering, statistical modeling, finance, and machine learning.

We embrace the open-source philosophy, and so this course contents and the software/datasets used are open to the public under a permissive license.

Requirements

Prospectus students must:

• Have their own hardware.
  • At least 4GB of RAM. Recommended: 8GB or more.
  • Access to the internet.
• Have basic knowledge about the following topics.
  • Finance (e.g. interest rates, cash-flow analysis, profiling)
  • Mathematics (e.g. linear-algebra, calculus, probability & statistics)
  • Computer Science (e.g. programming languages, data structures, algorithms).

What you will learn

Successful completion of this course should enable students to:

• Understand the financial theory behind credit risk.
• Leverage different data-systems involved in the credit analysis process.
• Create credit models using machine learning techniques.
• Get exposure to the most relevant big-data technologies on the financial industry.

Syllabus

This course covers three main topics:

1. Financial concepts & tools behind credit scoring.
2. Data processing and management in financial applications.
3. Credit risk modeling (predictive models) and scoring.

Grading

The grading system is fairly simple:

Where n and m are undefined integers greater than 5.

Homework (40%)

Consider the following homework types:

• Data Entry

  Student answering surveys or uploading a particular set of data points. Read the privacy notice for more information.

• Exercise

  Math or programming exercises to demonstrate technical proficiency.

• Report

  Technical document containing detailed analysis on a subject. Technical proficiency, analysis, and communication skills are evaluated.

• Essay

  Opinion-based writing with in-depth analysis and insight on a particular subject.

• Code

  Functional application following the coding best practices. Includes documentation and testing.

Projects (20%)

There will be two projects. Final dates are to be defined, but tentatively:

• Project 1: W10/S02 (10%)
• Project 2: W16/S01 (10%)

Exams (30%)

There will be three exams; one per general topic. Final dates are to be defined, but tentatively:

• Exam 1: W06/S01 (10%)
• Exam 2: W11/S02 (10%)
• Exam 3: W16/S01 (10%)

Values

• Excellence
• Collaboration
• Transparency (honesty)
• Open-mindedness (continuous learning)

Attendance monitoring

Student attendance will not be enforced by the professor. To gain the “attendance point of the day” the student must sign the attendance monitoring sheet during the first 10 minutes of the lecture.