Brute force - This method most inefficient of all as there is more possibilty for the robber to get caught. So, we can eliminate this.
Divide and Conquer - He can only visit some of the houses and not all. So we can eliminate this.
Dynamic programming - This is better than above two but while identifying subproblem,he would decide to rob alternate houses(He wouldn't get caught). But by doing so he wont be able to get maximum profit. This is good algorithm because he can start from house where highest amount of money is stored.
Backtracking - The robber can rob at alternate houses and at the end of street he can backtrack all the way back to start of the street and again start robbing the houses he had skipped in the first round. By doing this way he might steal all the houses in street. But this doesn't always happen always, backtracking would select any one of the choices currently available if it provides the solution then it continues else it backtracks and chooses another choice. Given the circumtances, robber cannot do trails(he might get caught).
So, Option B