Optimal gradient-based learning using importance weights.

Sepp Hochreiter, Klaus Obermayer

Browse the full IJCNN paper archive.