Suppose you have access to a large number of game logs for players participating in the game The Resistance. This is a bluffing game where players are either spies or true members of the resistance, and a large part of the game is finding who are the spies. The game logs you have are based on the voting record of each player, and includes: what round the player voted in, how many missions had already failed (the score), whether they were on the team, whether they voted for or against the mission, what was the mission outcome, (success, fail, or not approved) and whether they were ultimately revealed to be a spy. A sample of the data is in Table 1: Round 3 2 1 2 3 1 2 3 3 3 2 2 Score Team Vote 1 No For 1 No Against 0 No Against 0 Yes Against Not Approved 2 Yes For 0 Yes For 1 No Against For 2 No Yes No 0 1 1 1 No Yes Against Against For For Outcome Success Success Fail Fail Success Not Approved Not Approved Success Not Approved Fail Not Approved Spy Yes No Yes No Yes No Yes Yes Yes Yes No No Table 1: Voting behaviour of players in The Resistance, and whether they were ultimately revealed to be a spy. Describe the process you would use to induce a decision tree to help determine whether a player is a spy based on their voting behaviour. . Given the data in Table 1, which attribute would you use to branch on first and why?