Hi Don Li,

Decision trees can be used to solve supervised learning problems. They can be especially useful when the relationship between the inputs and outputs are not linear. This chapter from Christoph Molnar's book Interpretable Machine Learning covers decision trees really well.