A. Marcu

We propose FMix, an MSDA that uses binary masks obtained by applying a threshold to low frequency images sampled from Fourier space, obtaining new state-of-the-art results on CIFAR-10 and Fashion-MNIST. Expand

We analyse MixUp, CutMix, and FMix from an information theoretic perspective, characterising learned models in terms of how they progressively compress the input with depth. Expand

In this paper, we consider an online scheduling problem where a number of agents compete for a perishable resource and a fair allocation needs to be found. Expand

Data supporting the paper: E. Gaspers, A. Mattei, T. Marcu, N. Perez-Diaz, H. Aziz, S. Walsh. Expand

In this paper, a new approach to computing the generalisation performance is presented that assumes the distribution of risks, $\rho(r)$, for a learning scenario is known. Expand