Course 2025-2026 a.y.

21046 - DATA SCIENCE AND MACHINE LEARNING FOR FINANCE

Department of Finance

Course taught in English
31
ACME (6 credits - II sem. - OP  |  SECS-P/11) - AFC (6 credits - II sem. - OP  |  SECS-P/11) - AI (6 credits - II sem. - OP  |  SECS-P/11) - CLELI (6 credits - II sem. - OP  |  SECS-P/11) - CLMG (6 credits - II sem. - OP  |  SECS-P/11) - DES-ESS (6 credits - II sem. - OP  |  SECS-P/11) - EMIT (6 credits - II sem. - OP  |  SECS-P/11) - ESS (6 credits - II sem. - OP  |  SECS-P/11) - FIN (6 credits - II sem. - OP  |  SECS-P/11) - GIO (6 credits - II sem. - OP  |  SECS-P/11) - IM (6 credits - II sem. - OP  |  SECS-P/11) - MM (6 credits - II sem. - OP  |  SECS-P/11) - PPA (6 credits - II sem. - OP  |  SECS-P/11)
Course Director:
FRANCESCO CORIELLI

Classes: 31 (II sem.)
Instructors:
Class 31: FRANCESCO CORIELLI


Suggested background knowledge

This course is based on a mix of Statistics/Econometrics and coding. The following abilities would be very useful: Knowledge of Statistics beyond the basic Statistics courses (e.g. a course in Econometrics) Knowledge of matrix notation and elementary matrix algebra Some knowledge of Python possibly as implemented in Google Collaborate A general knowledge of financial markets, in particular the stock market and bond market could be useful (however, the necessary details shall be presented during the course)

Mission & Content Summary

MISSION

This course is designed to introduce students to the applications of machine learning (ML) in finance. The course shall cover the fundamental concepts of ML, including supervised and unsupervised learning, neural networks, deep learning, and natural language processing. Students will learn how to apply these concepts to financial data and build predictive models to support financial decision-making.

CONTENT SUMMARY

  1. Introduction: data science and statistical methods in Finance
  2. How to use Python in the Colaboratory environment
  3. List of models studied in the course
  4. Summary of some classical method (Regression, Principal Components, Canonical Correlation, Cluster analysis)
  5. Generalization of these models in Machine Learning
  6. Introduction to Neural Networks (NN)
  7. Overview of NN models
  8. Focus on autoencoders
  9. Autoencoders and factor models
  10. Autoencoders for pricing, hedging and trading
  11. Introducing time dependency: Recurrent Neural Networks
  12. Introducing context dependency: LLM and the Attention Mechanism

Intended Learning Outcomes (ILO)

KNOWLEDGE AND UNDERSTANDING

At the end of the course student will be able to...

 

  1. Understand what is machine learning 
  2. Understand how ML method originate in classical statistical methods and in which directions evolved from these
  3. Understand in detail of how neural networks work in deep learning
  4. Create models based on autoencoders, recurrend NN, attention mechanisms and LLM
  5. Apply these methods in finance, in particular in pricing, hedging and trading

APPLYING KNOWLEDGE AND UNDERSTANDING

At the end of the course student will be able to...
  1. Use Colab and Python to create and test ML models
  2. Apply autoencoders to financial data for pricing, hedging and trading
  3. Build a DL model using sequential, recurrent and attention layers
  4. Understand the properties and potential problems of a predefined ML model
  5. Avoid the most common pitfalls in ML and DL

Teaching methods

  • Lectures
  • Practical Exercises
  • Collaborative Works / Assignments

DETAILS

Lectures shall be face to face

Al lectures are based on Python programs implemented as Jupyter notebooks in Google Colaboratory

During lectures, students shall be able to work on the same programs as the teacher


Assessment methods

  Continuous assessment Partial exams General exam
  • Written individual exam (traditional/online)
     x
  • Collaborative Works / Assignment (report, exercise, presentation, project work etc.)
 x    

ATTENDING AND NOT ATTENDING STUDENTS

Optional Takehome (group) assignement + Written final exam

The final exam shall provide further questions for those who decide not to do the Takehome

For those who do the takehome, 80% of the final grade comes from the written exam, 20% from the takehome

For those who choose not to do the takehome100% of the grade comes from the exam with added questions (as described above) 

Examples of questions for the final written exam shall be provided during the course

Questions in the written exam and in the Takehome shall cover the ILOs of the course


Teaching materials


ATTENDING AND NOT ATTENDING STUDENTS

Colab workbooks, slides and handouts provided by the teacher and available in BBoard.

 

Optional accompanying textbooks:

-Auréliene Géron, "Hands-On Machine Learning with Scikit-Learn Keras & tensorFlow", 3rd edition (2022), O°Reilly Media Inc.

--Book Github: GitHub - ageron/handson-ml3: A series of Jupyter notebooks that walk you through the fundamentals of Machine Learning and Deep Learning in Python using Scikit-Learn, Keras and TensorFlow 2.

-AA. VV., "Dive Into Deep Learning", 1st edition (2023), C.U.P. 

--Book Website: Dive into Deep Learning — Dive into Deep Learning 1.0.3 documentation

Last change 21/05/2025 10:48