Machine Learning in Finance
Objectives
Machine learning is a discipline within the field of Artificial Intelligence, which, through algorithms and statistics, provides computers the ability to identify patterns in data and make predictions. This curricular unit will familiarize students with the most common machine learning algorithms and their applications in finance.
General characterization
Code
400103
Credits
7.5
Responsible teacher
Nuno Miguel da Conceição António
Hours
Weekly - Available soon
Total - Available soon
Teaching language
Portuguese. If there are Erasmus students, classes will be taught in English
Prerequisites
It is recommended that students have basic knowledge of statistics and Python.
Bibliography
Teaching method
The curricular unit is based on theoretical-practical classes. The sessions include the presentation of concepts and methodologies and the practical application of different concepts using different languages and computer applications. Several teaching strategies are applied, including slides presentation and step-by-step instructions on approaching practical examples, questions, and answers. The practical component is oriented towards exploring tools introduced to students, including the discussion of the best approach in different scenarios.
Applications used: Python, Jupyter notebook, Microsoft visual code, Azure Machine Learning Studio (classic).
All submissions should be made via Moodle. Submissions after the deadline will be rejected.
Evaluation method
Due to the application-based design of the course, evaluation is continuous.
All evaluation grades are on a scale of 0-20. The final course grade is calculated based on the following weights:
- Completion of self-assessment survey: 2.5%
- Group membership submission (in the due deadline): 2.5%
- Group project (minimum grade is 8.0):
- Materials (datasets, code, etc.) and report: 55%
- Due date: to be confirmed (around 7 days before 1st season exam date)
- Exam (individual - with materials consultation - minimum grade is 8.0): 40%
Subject matter
- LU1. Introduction to Machine Learning, Data Mining, and Predictive Analytics.
- LU2. Machine Learning applications in finance.
- LU3. Machine Learning libraries for Python and introduction to AZMLS.
- LU4. CRoss-Industry Standard Process for Data Mining (CRISP-DM) Methodology.
- LU5. Data understanding.
- LU6. Data preparation.
- LU7. Model validation, generalization, and overfitting.
- LU8. Supervised learning - regression: performance measures.
- LU9. Main families of algorithms: linear regression, decision trees, neural networks, Support Vector. Machines (SVM), K-Nearest Neighbors (KNN).
- LU10. Supervised learning - classification: performance measures.
- LU11. Main families of algorithms: logistic regression, decision trees, neural networks, Naive Bayes, SVM, and KNN.
- LU12. Ensembles of methods.
- LU13. Unsupervised learning - clustering: K-Means algorithm.
- LU14. Application in finance cases.