What statements are true?

The goal count heuristic counts the number of satisfied goals

The goal count heuristic is rather uninformed because it ignores important aspects of the task

The optimal delete relaxation heuristic is ''easy'' to compute

There exist efficient algorithms that approximate the optimal delete relaxation heuristic

An optimal delete relaxation heuristic can overestimate the true cost of reaching the goal

The A* search algorithm with the optimal delete relaxation heuristic yields optimal plans

Consider a Blocksworld STRIPS planning task $\Pi = \langle V, I, G, A\rangle$ where $V$ and $A$ is given as in the chapter on STRIPS, initial state $I$, and goal $G$ are as follows:

• $I = \{\textit{clear}_B, \textit{on}_{B,G}, \textit{on}_{G,R}, \textit{on-table}_R \}$
  
• $G = \{\textit{clear}_R, \textit{on}_{R,B}, \textit{on}_{B,G}, \textit{on-table}_G \}$

What is the heuristic estimate $h^+(I)$?