|
Bart Goossens |
- Security of neural networks
- Robustness against adversarial perturbations
|
|
- Fuzzy logic
- Rule-based modelling
|
|
- Fuzzy logic
- Rule-based modelling
|
|
- actionable AI for systems with Human in the Loop
|
|
|
- Fusion of Semantics & ML
- Provenance
- Reasoning
|
|
- Knowledge graph embeddings (to fuse domain knowledge into machine learning)
- Association rule mining
- Inductive Logic Programming
- Rule extraction/learning
- White-box machine learning
- Genetic algorithms
|
|
|
|
|
|
|
|
- Counterfactual explanations
- Metaheuristics
|
|
|
|
- Explainable network embeddings
|
|
- Feature extraction and visualization
- Trustability and robustness (adversariality)
|
|
- Explainable deep learning
|
|
