Hi! I’m Albert, a PhD Candidate at Delft University of Technology, specializing in Tensor Network Machine Learning advised by Prof. Dr. ir. Kim Batselier.

My research is focused on probabilistic modeling, tensor networks, and machine learning. I develop methods that make AI systems more efficient, interpretable, and uncertainty-aware by combining Bayesian approaches with tensor networks (CP, Tucker, TT) in deep and kernel-based architectures.

I’m particularly interested in building scalable and reliable learning systems that retain expressiveness while reducing computational and memory costs. My work has applications in recommender systems, knowledge graphs, and privacy-preserving federated learning.

Beyond research, I enjoy translating theory into practice — applying ML methods to real-world data and large-scale systems. I work across the full ML pipeline: data processing (SQL, Pandas, PySpark), scientific computing (NumPy, SciPy), and model development (Python, PyTorch, JAX, scikit-learn, and LLM-based technologies).

Broadly, I’m driven by the goal of creating efficient and trustworthy AI — systems that are lighter, faster, and more reliable through principled mathematical design.