Disadvantages of decision trees

The model stops only when all data points can fit into a single class/category. So there is a possibility that it may not generalize well for complex problems and the chance of bias is high.

These problems can be solved by defining the maximum depth of the tree or by specifying the minimum number of data points needed to split the node further in the tree.