Methods for Learning Interpretable, Reliable and Robust Deep Models
Description
The research project aims to develop deep neural network learning methods to improve the interpretability, reliability and robustness of the models produced. The proposed approach is based on the extraction of symbolic components manipulating conceptual abstractions to describe the reasoning used by the neural network leading to decision making for a specific context, in addition to the decision itself. More specifically, the PhD project focuses on the development of methods for building concepts into the neural network structure from training data. The said concepts will be created to be used as variables for the description of a reasoning leading to a decision in a symbolic form (e.g. rules, computer program). This work should lead to the proposal of new methods for better explaining complex model decisions by domain experts, while enabling more formal verification of reasoning by algorithmic means.
Research Field
- Machine Learning
- Deep learning
- Artificial Neural Networks
Research Supervisor
Christian Gagné
Research Environment
Institute Intelligence and Data (IID) of Université Laval
The IID is dedicated to federating and supporting expertise and innovation in artificial intelligence and data valorization in the greater Quebec City area, placing itself at the heart of a global centre of excellence of international calibre in these fields and related research areas.
Web Site
Financial Aid Available by Program of Study
Doctorate in Computer Science
Program descriptionFinancial Aid Available*
Financial Aid Related to Research Project
$30000 per year for 4 years.
Program-Specific Financial Aid
Graduate Studies Awards
| Milestone |
Amount |
Progression scholarship 1 - 7
|
7 x $1,600 |
| Progression scholarship 8 |
$800 |
| Total |
$12,000 |
Université Laval: Student Financial Aid
Supplemental Tuition Fee Exemption Scholarship Program: Entitles international students to pay Canadian student tuition fees, for overall savings of around $49,000.
* Amounts shown represent maximum financial aid available. Certain conditions apply. Subject to change without prior notice. For further information, contact sponsoring organizations directly. Doctorate in Electrical Engineering
Program descriptionFinancial Aid Available*
Financial Aid Related to Research Project
$30000 per year for 4 years.
Program-Specific Financial Aid
Graduate Studies Awards
| Milestone |
Amount |
Progression scholarship 1 - 7
|
7 x $1,600 |
| Progression scholarship 8 |
$800 |
| Total |
$12,000 |
Université Laval: Student Financial Aid
Supplemental Tuition Fee Exemption Scholarship Program: Entitles international students to pay Canadian student tuition fees, for overall savings of around $49,000.
* Amounts shown represent maximum financial aid available. Certain conditions apply. Subject to change without prior notice. For further information, contact sponsoring organizations directly.
Desired Profile
- Computer Software
- Electrical Engineering
- Computer Engineering
- Software Engineering
- Data science
- Artificial intelligence
Requirements and Conditions
Applicants for this project must demonstrate a willingness to take part in an ambitious computer science research project, with in-depth knowledge of machine learning and deep learning, and a desire to establish themselves in the international academic research community.
Required Documentation
- Curriculum vitæ
- Student transcript
Transmit the scientific publications or other productions demonstrating research experience.
Find Out More
Christian Gagné
Professeur
Département de génie électrique et de génie informatique
christian.gagne@gel.ulaval.ca