Fetching the paper…

Enhanced Expressive Power and Fast Training of Neural Networks by Random Projections · Around