Aku Kammonen is a Research Scientist in the Stochastic Numerics Research Group under the supervision of Professor Raul F. Tempone at King Abdullah University of Science and Technology (KAUST).

Research Interests

Aku's research interests include Machine Learning, Neural Networks, Random Features, Spectral Bias, Adversarial Attacks, and Numerical Analysis.

Selected Publications

• A. Kammonen, L. Liang, A. Pandey, and R. Tempone (2024, March 19–23). Comparing Spectral Bias and Robustness For Two-Layer Neural Networks: SGD vs Adaptive Random Fourier Features [Poster presentation], International Conference on Scientific Computing and Machine Learning 2024, Kyoto, Japan
• A. Kammonen, J. Kiessling, P. Plecháč, M. Sandberg, A. Szepessy, R. Tempone, Smaller generalization error derived for a deep residual neural network compared with shallow networks, IMA Journal of Numerical Analysis, Volume 43, Issue 5, September 2023, Pages 2585–2632
• A. Kammonen, J. Kiessling, P. Plecháč, M. Sandberg, A. Szepessy, Adaptive random Fourier features with Metropolis sampling, Foundations of Data Science, 2, 3, 309, 2020
• A. Kammonen, P. Plecháč, M. Sandberg, A. Szepessy, Canonical Quantum Observables for Molecular Systems Approximated by Ab Initio Molecular Dynamics, Ann. Henri Poincaré 19, 2727–2781, 2018
• A. Kammonen, On a one-phase quasi-static Stefan problem for planar polygonal crystals grown from vapor in a bounded container, Advances in Mathematical Sciences and Applications Vol. 24, No.2 (2014), pp.317–331

Education Profile

• Ph.D. Applied and Computational Mathematics specialized in numerical analysis, Royal Institute of Technology (KTH), Sweden, 2021
• M.Sc. in Engineering, Microelectronics, International track, Japanese, Sweden, 2014