A5043575212- Model Reduction and Neural Networks
- Natural Language Processing Techniques
- Speech and dialogue systems
- Probabilistic and Robust Engineering Design
- Topic Modeling
- Neural Networks and Reservoir Computing
- Neural Networks and Applications
- Neural dynamics and brain function
- Advanced Memory and Neural Computing
- Iterative Methods for Nonlinear Equations
- Fluid Dynamics and Turbulent Flows
Monash University
2023-2024
Australian Regenerative Medicine Institute
2023
Physics-informed neural networks (PINNs) and their variants have recently emerged as alternatives to traditional partial differential equation (PDE) solvers, but little literature has focused on devising accurate numerical integration methods for (NNs), which is essential getting solutions. In this work, we propose adaptive quadratures the of apply them loss functions appearing in low-dimensional PDE discretisations. We show that at opposite ends spectrum, continuous piecewise linear (CPWL)...
Neuromorphic computing aims to emulate biological neural functions overcome the memory bottleneck challenges with current Von Neumann paradigm by enabling efficient and low-power computations.In recent years, there has been a tremendous engineering effort bring neuromorphic for processing at edge.Oscillatory Neural Networks (ONNs) are braininspired networks made of oscillators mimic neuronal brain waves, typically visible on Electroencephalograms (EEG).ONNs provide massive parallelism using...
Physics-informed neural networks (PINNs) and their variants have recently emerged as alternatives to traditional partial differential equation (PDE) solvers, but little literature has focused on devising accurate numerical integration methods for (NNs), which is essential getting solutions. In this work, we propose adaptive quadratures the of apply them loss functions appearing in low-dimensional PDE discretisations. We show that at opposite ends spectrum, continuous piecewise linear (CPWL)...
Physics-informed neural networks (PINNs) and their variants have recently emerged as alternatives to traditional partial differential equation (PDE) solvers, but little literature has focused on devising accurate numerical integration methods for (NNs), which is essential getting solutions. In this work, we propose adaptive quadratures the of apply them loss functions appearing in low-dimensional PDE discretisations. We show that at opposite ends spectrum, continuous piecewise linear (CPWL)...