Titulo Estágio
Financial Interpretability with Intelligent Methods
Áreas de especialidade
Sistemas Inteligentes
Local do Estágio
CISUC
Enquadramento
Deep learning methods, as convolutional and recurrent neural networks, are becoming standard go-to algorithms for a wide range of applications. However, applicability in several critical applications, e.g. public policy, security/safety systems, health diagnosis and fraud detection, has been faced with some hurdles do to lack of model interpretability.
Interpretability has been a focus of research since the beginning of Deep Learning, because high accuracy and high abstraction bring the back box problem, i.e. accuracy vs interpretability problem. This aspect is also of importance because of trustworthiness issues, i.e. a model that is not trusted is a model that will not be used. This issues often arise in real application scenarios, where end-users are not easily convinced on the reliability of black box model.
In this internship it is proposed to research and implement interpretability mechanisms that can be applied in deep learning models applied to financial applications. The context will be a European project towards a transparent financial industry.
Objetivo
In this internship the student should study, propose, implement, and test methods for interpretability and trustworthiness in deep learning models.
To achieve this goal, the following objectives will be pursued:
- Study the state of the art
- Study the available frameworks for model development
- Define case studies for financial applications
- Define, implement, and fine tune the interpretability architecture
- Propose and deploy test setup
Plano de Trabalhos - Semestre 1
- Literature review
- Identification and study of interpretability mechanisms
- Identification and study of available frameworks
- Analyse and define case studies
- Define the architecture of the system
- Start implementing the proposed approach
- Write intermediate report
Plano de Trabalhos - Semestre 2
- Implement the proposed solution and fine tune models
- Test and evaluate performance
- Write final report
Condições
This work should take place in the context of a CISUC lab. This work is part of a European Consortium allowing teaming up with an international network.
Orientador
Bernardete Martins Ribeiro e Catarina Helena Branco Simões da Silva
bribeiro@dei.uc.pt 📩