New beam search algorithms optimize scheduling, reducing tardiness and costs significantly.
The article discusses ways to improve scheduling on a single machine by reducing delays and setup times. The researchers used beam search algorithms to find the best solutions. They developed new versions with variable widths that outperformed the fixed width ones. The filtered beam search algorithms gave the best results, but took longer for large problems. The priority beam search algorithms were faster and suitable for bigger instances. The 3-swap method was the most effective in reducing costs for many cases.
