معرفی
Professor Barak Pearlmutter is affiliated with Maynooth University in the Faculty of Science & Engineering. His research spans multiple domains including automatic differentiation, neural networks, machine learning, and neuroscience. He has contributed significantly to adaptive systems, brain imaging, and programming language design.
- Research Interests include: Adaptive systems, automatic differentiation, theoretical neurobiology, neural networks, machine learning, acoustic source separation/localization, neuroscience, brain imaging, programming language design, and computational neuroscience.
- Publications focus on applying algorithmic differentiation to machine learning, developing neural ODE models for biomedical signals, advancing sparse NMF techniques, and integrating functional programming with numerical methods.
- Collaborations span institutions like MIT, Oxford, and IEEE societies, with work in brain-computer interfaces, MEG source localization, and neuromodulation for tinnitus treatment.
Technical Contributions include the DiffSharp AD library for .NET languages and foundational work on reverse-mode automatic differentiation in functional frameworks. His 2018 Journal of Machine Learning Research survey on AD remains a seminal reference in the field.
Application Areas cover biomedical signal processing, optical brain-computer interfaces, cognitive modeling, and neural code optimization. His work intersects computer science, neuroscience, and mathematical computing through sparse decomposition and probabilistic modeling.
