HIGHLIGHTS
- who: Adrian Barbu from the (UNIVERSITY) have published the paper: Training a Two-Layer ReLU Network Analytically, in the Journal: Sensors 2023, 23, 4072. of /2023/
- what: The authors will explore an algorithm for training two-layer neural networks with ReLU-like activation and the square loss that alternatively finds the critical points of the loss function analytically for one layer while keeping the other layer and the neuron activation pattern fixed. Even though the scope is narrow, the authors show that in some cases, the optimization capabilities of the proposed method greatly outperform those . . .
If you want to have access to all the content you need to log in!
Thanks :)
If you don't have an account, you can create one here.