Estimating effective connectivity in neural networks: comparison of derivative-based and correlation-based methods
Niklas Laasch, Wilhelm Braun, Lisa Knoff, Jan Bielecki, Claus Christian Hilgetag
The paper compares methods for estimating effective connectivity in neural networks, finding that lagged-cross-correlation performs best in sparse networks with delays and noise.
bioRxiv · Feb 5, 2024 · PreprintNeuroscience