Automated algorithm selection using meta-learning and pre-trained deep convolution neural networks

Itai Dagan, Roman Vainshtein, Gilad Katz, Lior Rokach

Abstract: Highlights•A zero-shot approach for selecting classification algorithms for tabular datasets.•A novel approach for modeling whole datasets as images, which enables cross-dataset learning.•Our model achieves results comparable to state-of-the-art solutions, at a fraction of the computational cost.