What are some examples that cause Simple Hill Climbing to reach problems like local maxima, ridges and alleys, and plateau problem(s)? I have tried searching:
- Link one: which gives a fairly good example of Simple Hill Climbing stuck-in-local-maxima problem in block arrangements. However, it does not show the steps.
- Link two: which gives steps to finding the solution in SHC. However, I don't understand how h(1) can be -6 when all there are only four blocks and four of them are misplaced, therefore only yielding -4. It also doesn't show the problem(s) encountered by SHC.
- Link three: I understand how the concept of reaching state 'g' makes your algorithm reach a local maximum and get stuck. However, it's rather ambiguous what the states are and I don't know what state 'g' and the final states refer to.
From the lecture note I read, I was given the TSP problem. The graph was a complete graph with four nodes: A, B, C and D. We used both Simple Hill Climbing and Steepest-Ascent Hill Climbing to solve the problem. The heuristic value used to solve that problem was the total distance of each state. We can explore other neighbouring states by switching positions of the characters "ABCD" using 6 different combinations (first letter <-> second, second <-> third, etc.). However, in the example given, it did not show what exactly "stuck at a local maximum" is. Neither did it show the ridges and alleys problem nor the plateau problem.
Could someone give me an example of how we reach those problems and what those problems actually are in examples (I understand the definition of each problem: here and here)? For reference, this below is the image of the TSP problem I mentioned: