Enhancing Zero-shot Chain of Thought Prompting via Uncertainty-Guided Strategy Selection