I understand how Tabu search works, namely how it is similiar to Hill Climbing, however sets of points in the search space are remembered. This is called the Tabu list as the algorithm tries to avoid them.
I then came across this quote which can be either true or false:
"It uses a memory data structure to prevent moves to previously visited points in the search space."
This seems right... How does Tabu search use 'memory data structure'? I know it uses memory structure, but memory data structure seems wrong. Am I thinking too much into this, or am I right in being weary because data structures could mean something else entirely.
It needs to remember what is tabu, so it needs to store that in memory in a data structure. For example I use a LinkedList.
Note: Making "visited points in the search space" tabu doesn't scale. That's Solution tabu. Other forms of tabu, such as Move tabu (not doing the same move again), Entity tabu (not moving the same thing again) and Value Tabu (not moving to the same value again) work much better and scale much better.