Discovering Heuristics with Large Language Models (LLMs) for Mixed-Integer Programs: Single-Machine Scheduling

A recent study has made significant strides in the field of scheduling and combinatorial optimization by utilizing Large Language Models (LLMs) to discover new heuristics. This research specifically addresses the single-machine total tardiness problem, which is crucial for improving efficiency in job sequencing on a single processor. By developing and benchmarking two innovative heuristics, the study not only enhances our understanding of optimization techniques but also opens up new avenues for practical applications in various industries, making it a noteworthy advancement in the field.