Various modifications of TRANSFORMER were recently used to solve time-series forecasting problem. We propose Query Selector - an efficient, deterministic algorithm for sparse attention matrix. Experiments show it achieves state-of-the art results on ETT, Helpdesk and BPI'12 datasets.
翻译:最近对TRRANSORMER的各种修改被用于解决时间序列预测问题。 我们提议了查询选择器 — 一种高效的、确定性算法, 用于稀疏关注矩阵。 实验显示它取得了在 ETT、 服务台和 BPI' 12 数据集上的最新结果 。