Long-term series forecasting with Query Selector -- efficient model of sparse attention

This addresses the challenge of efficient long-term forecasting for time-series analysis, likely incremental as it builds on existing Transformer modifications.

The authors tackled the problem of long-term time-series forecasting by proposing Query Selector, an efficient deterministic algorithm for sparse attention in Transformers, achieving state-of-the-art results on ETT, Helpdesk, and BPI'12 datasets.