Discussion 1

learning and stable matching