Text this: Traffic Signal Control: a Double Q-learning Approach