Oskar Allerbo

Oskar Allerbo is a postdoc researcher at the Royal Institute of Technology (KTH), affiliated with the Probability, Mathematical Physics and Statistics Unit. His primary research interests span statistics, machine learning, neural networks, and regularization methods, with a focus on theoretical and applied aspects.

• Institution: Royal Institute of Technology (KTH)
• Role: Researcher (postdoc)
• Research Areas: Statistics, Machine Learning, Regularization, Neural Networks, Medical Technology

His recent work explores robust kernel regression, elastic net solution paths, and theoretical properties of neural networks. He also investigates applications in medical technology, particularly anti-inflammatory treatments through vagus nerve stimulation. Methodologically, he connects mathematical statistics with biomedical engineering problems.

Key publication trends include: (1) kernel methods and gradient descent optimization, (2) sparse regression and autoencoder architectures, (3) regularization theory in neural networks, and (4) interdisciplinary applications in biophysics and medical technology. Manuscripts in progress address double descent phenomena, bandwidth selection, and generalization mechanisms.