Introduction - If you have any usage issues, please Google them yourself
Random forests are used in a random way to build a forest, there are a lot of decision trees in the forest, there is no association between the random forest of each decision tree. In the forest, when there is a new input into the sample, let every decision tree in the forest was a judge, look at the sample which should belongs to a class, and then take a look at what kind of choice is the most, the predicted samples for that class.