Current Solution

The next question which we should ask while framing the problem is that how are we solving the problem currently and how well the current solution is performing?

Suppose, on doing some research we find out that currently, the house prices are predicted manually by using some complex rules. It is costly and time-consuming and their estimates have 15% error.

So we get to know that our model should have less than 15% of error to replace the current solution.