What is a decision tree ?

A decision tree is a tool that uses a tree-like model to analyze decisions and their possible consequences.

Decision trees are usually used in decision analysis, to help identify a plan of action that is most likely to reach a goal.

A decision tree is very similar to flow charts because it is used to envisage the decision making process by mapping out different courses of action, as well as their potential outcomes.

It is drawn from left to right and has different paths that split off but never converge.

Decision trees are usually drawn using a pen and paper but specialized software can also be used.

Decision trees usually consist of three different elements, which are :

Root Node: This is used to represent the main decision that you are trying to make.

Branch node : these are used to represent different courses of action that are available when making a particular decision.

Leaf node : this is used to represent possible outcomes for each action. The leaves are nodes that are attached at the end of the branches.

There are two types of leaf nodes: square leaf nodes, which represent additional decisions to be made, and circle leaf nodes, which represent unknown outcomes.

Why are decision trees used ?

Decision trees are used because they have numerous advantages, which include the following :

  1. Decision trees are flexible: Decision trees are used to easily plan, predict and analyze the possible outcomes to your decisions.

For example, a stakeholder can use a decision tree to decide between buying a software or leasing one.

Some of the criteria that can be used to help make that decision include the total cost of ownership, regulatory implications, importance of application to the organization and return on investment.

2. Decision trees can be used to understand complex processes : Decision trees can be used to represent cause and effect relationships, and simplify a complex process.

So how do you draw a decision tree?

  1. Start with the main decision : The main decision should be at the first thing that you draw in your decision tree. This is the root of the entire decision tree diagram.
  2. Draw branches to connect the possible course of actions : write each of the possible courses of action below the main decision i.e. the root and connect them to the root using a straight line. The costs of each of these possible course of action should be written on the lines along with the probability for success.
  3. Draw leaf nodes at the end of your branches : the result of each course of action should be written at the end of the branches in the form of leaves. If a different decision is to be made, draw a square leaf node but if the outcome is uncertain, draw a circular leaf node.
  4. Determine the likelihood of success of each decision point: use carefully researched data to predict the likelihood for success.
  5. Evaluate risks versus the rewards of each decision: you should determine the expected value of each decision in tree to help minimize the risk and help increase the likelihood of success.

In this example below, a company is using a decision tree to decide on how to invest their money.

The available options include stocks, mutual funds and bonds.