Home

AI4Industry Legal and Financial (AI4ILF) @ IPLAB

The AI4Industry Legal and Financial (AI4ILF) is a cutting-edge research area dedicated to the development and application of bio-inspired models for deep learning in industrial, legal, and financial/commercial domains. The reaserch area’s mission is to advance artificial intelligence (AI) methodologies by leveraging bio-inspired computational models, innovative mathematical approaches, and novel learning and testing procedures to build AI-driven solutions. In the industrial field, the research focuses on automotive applications (e.g., driver assistance, autonomous driving, engine control, powertrain optimization) and semiconductor technologies (e.g., production line optimization, semiconductor modeling, CFD analysis, AI-enhanced SPICE models, residual lifetime prediction etc...). Concurrently, the research area explores legal and financial AI applications, with a special emphasis on civil, corporate, and contract law, as well as financial risk assessment, insolvency prediction, AI-based company valuation, and advanced time-series forecasting. The research extends across advanced AI paradigms, including convolutional neural networks (CNNs), transformers, hybrid systems, reinforcement learning (RL), and innovative learning architectures. Key topics of interest include, but are not limited to:

  1. Perceptive deep learning for industrial applications;
  2. Adversarial methods, regularization techniques, and adaptive Jacobian regularization;
  3. Continual learning and mathematical analysis of learning dynamics, addressing catastrophic forgetting via Lipschitz properties, Extended Elastic Weight Consolidation (EWC), and bio-inspired models;
  4. Learning in non-conventional spaces and geometries, particularly hyperbolic spaces;
  5. Development frameworks based on perceptive deep learning and generative models;
  6. Generative AI for legal applications;
  7. Domain adaptation for automotive applications, including multi-scaled gradient reversal layers and generative layers for domain adaptation;
  8. Explainability through foundation models and Cellular Neural Networks (CNNs);
  9. Knowledge distillation for resource-constrained AI systems;
  10. Neuro-modulation paradigms balancing stability and plasticity in deep networks;
  11. Document analysis using generative AI within RAG attention-based architectures;
  12. Hybrid bio-inspired models for financial data analysis;
  13. Temporal Fusion Transformer (TFT) with TCN enhancement for robust time-series forecasting;
  14. Attention-based models and explainability techniques for document analysis in civil and corporate law.

The AI4ILF research area is actively fostering strong scientific collaborations with renowned industry leaders, bridging fundamental AI research with real-world applications. Through these partnerships, the involved researchers aim to drive applied R&D, supporting the professional development of Ph.D. candidates, Ph.D students and so on.