I am looking for an algorithm to transform an input data to a goal data using a series of operations. The shorter the series the better.
The following is known:
- the input data
- the goal data
- input and goal data does not stand in any correlation
- operations (and there impact to the current data) which can be endless combined
- different input data for the same goal data could have same, similar or totally different operation series
- for some data states not all operations are possible
I thought of a pathfinding algorithm, since I can calculate the distance between current data and goal data. So each edge would be an operator and each node the current data. But I am unsure about variety and combination amount of possible operations.
What approach could I try?