Also known as dynamic optimization, DP
problem optimization method that simplifies a complicated problem by decomposing it into simpler subproblems recursively
via PubMed
Discovered by embedding cosine similarity (sentence-transformers MiniLM, 384-dim).