Modeling, Programming and Machine Learning for Financial Analysis


The student will have acquired knowledge, skills and competences that will allow him / her to :

- Know the main paradigms for decision support systems and machine learning applied to
financial data.
- Know the main concepts of programming and construction of decision support systems
applied to financial data.
- Be able to use and develop software for decision support and financial data analysis.
- Understand the basic types of neural networks, cost functions, the parameter optimization
method, the role of hyperparameters (number of layers, number of units in each layer and the
learning coefficient) and the various types of regularization.
- Be able to use software that implements neural networks to solve problems.

General characterization





Responsible teacher

Rui Alberto Pimenta Rodrigues


Weekly - 4

Total - 48

Teaching language





● Python for Finance: Financial modeling and quantitative analysis explained. Yuxing Yan.
Packt Publishing. 2017.
ISBN: 978-1783284375
● Tutorial: Sistemas de Apoio à Decisão sobre Indicadores Financeiros, Nuno Marques
● Deep Learning with Python, Francois Chollet,
Manning Publications Co. Greenwich, CT, USA ©2017
ISBN:1617294438 9781617294433

Teaching method

Concepts presentation, followed by examples and exercises, including computational
implementation. We will use real data.

Evaluation method

There will be two tests and a pratical work.  The final result is


Where T1 and T2 are the tests results and W is the pratical work note.

Subject matter

1) Decision Support Systems (DSS) in Finance: charts, characterization and models with Machine
Learning (ML).
2) Computation with financial data: data sources. Structure of a database for financial indicators and
indices (EUROSTOXX and S & P500). Algorithms: technical and fundamental analysis. Backtesting
and performance measures (alpha, beta and sharpe ratio).
3) Neural Networks:
Basic concepts: network types (feedforward, convolution and recurrent), activation and cost functions.
Parameter optimization and regularization.
non supervised learning: self-organizing maps (SOM).
4) DSS on financial data. Bachelier and Monte Carlo models. Self-similarity. Simulation and decision
with machine learning models.
5) Practical examples: developing and using relevant software.


Programs where the course is taught: