VALUE ITERATION ALGORITHM

How it works:

• This program performs the value iteration algorithm on some given model
• It infers that each state which only can transition back to itself is a terminal state
• After running, the program dislays the model used, the optimal value function and optimal policy
• Optionally, the sequence of steps to take according to optimal policy from start state to end state can be displayed (see below)

How to run:

• The command format should be: ./val_iter [input_file] [output_file] <optimal_policy_start_state> <optimal_policy_end_state>
• If the last 2 arguments are included then the sequence of optimal steps are displayed
• Run 'make' then 'make run' to run command: ./val_iter lab_data.txt output.txt 1 3

Input Data Format:

<< in a text file >> [number_of_states] [discount_factor] <state_1_transition_state> <state_1_transition_state_reward> ... <state_2_transition_state> <state_2_transition_state_reward> ... ...