dr. M. Shahsavari (Mahyar)
Universitair docent - Artificiële intelligentie
Universitair docent - Donders Institute for Brain, Cognition and Behaviour
Mahyar is an Assistant Professor at Artificial Intelligence (AI) group at Donders Institute for Brain, Cognition and Behavior. Mahyar previously was a research associate PostDoc at Imperial College London working on neuromorphic computing using parallel architectures, e.g., FPGAs. He was a PhD intern by Samsung at UK working on Machine learning deployment into android devices. He achieved a Master/Post-master at TUDelft and a PhD in Computer Science at CNRS, Lille University of Technology, where he developed a simulation platform for a Spiking Neural Network using the model of a memristor nanodevice as an artificial synapse.
He is interested in understanding the way brain operates and performs computation to mimic on computing machine. He worked on an artificial synapse that can remember and forget, which outperforms the artificial synapse that can only remember the last events. His research interests include neuromorphic computing, cognitive, brain-like and unconventional computing, machine learning applications running on novel architectures such as reconfigurable and parallel platforms. He likes swimming at nature, playing football and violin.
- Mahyar Shahsavari, Pierre Boulet, “Parameter Exploration to Improve Performance of Memristor-Based Neuromorphic Architectures”, IEEE Transactions on Multi-Scale Computing Systems, vol. 4, no.4, pp. 833 - 846, 2018. Volledige tekst
- Mahyar Shahsavari, Pierre Falez, Pierre Boulet, “Combining a Volatile and Nonvolatile Memristor in Artificial Synapse to Improve Learning in Spiking Neural Networks”, in IEEE/ACM International Symposium on Nanoscale Architectures (NANOARCH), Beijing, China, July, 2016. Volledige tekst
- Mahyar Shahsavari, Jonathan Beaumont, David Thomas, Andrew D. Brown, “POETS: A parallel cluster architecture for Spiking Neural Network ”, International Journal of Machine Learning and Computing (IJMLC), Vol.11(4), 281-285, 2021. Volledige tekst
- Mazdak Fatahi, Mahyar Shahsavari, Mahmood Ahmadi, Pierre Boulet, Arash Ahmadi, Philippe Devienne, “Rate-Coded DBN: An Online Strategy for Spike-Based Deep Belief Networks”, Journal of Biologically Inspired Cognitive Architectures, 2018. Volledige tekst