no code implementations • 10 Sep 2020 • Soha Sadat Mahdi, Nele Nauwelaers, Philip Joris, Giorgos Bouritsas, Shunwang Gong, Sergiy Bokhnyak, Susan Walsh, Mark D. Shriver, Michael Bronstein, Peter Claes, .
Since embeddings are accepted as an input, there is no need to train classifiers for the different properties and available data can be used more efficiently.