Trustworthy AI: A Computational Perspective