Multi-objective ω-Regular Reinforcement Learning