Mimicking Iterative Learning with Modern Hopfield Networks for Tabular Data
Sprache des Titels:
International Conference on Machine Learning (ICML 2022), 3rd Women in Machine Learning Un-Workshop
Hopular is a novel Deep Learning architecture for tabular data, where each layer is equipped with continuous modern Hopfield networks. Hopular is novel as it provides the original training
set and the original input at each of its layers. Therefore, Hopular refines the current prediction at every layer by re-accessing the original training set like standard iterative learning algorithms.