Decision Trees

NotesQuizObjective Test

Decision Tree

A Decision Tree is a diagram that looks at alternative courses of action and their possible outcomes

There are 2 stages to using Decision Trees:

  1. Draw the Decision Tree (including all probabilities and outcomes)

  2. Use expected values at all outcome points to make decisions.

Drawing the Decision Tree

  • Draw the tree from left to right

  • A square represents a Decision

  • A circle represents an Outcome

  • At a Decision Square - a branch from it represents a potential event - with a probability of it happening attached

Figure 1:

There are two branches coming off the initial decision point - the top branch has a certain outcome

The lower branch has two possible outcomes, and a further 2 possible outcomes for each of these

The next step would be to label the tree and put the cash inflows/outflows and probabilities in

Evaluation of the Tree

Evaluate the tree from right to left

  • Calculate an Expected Value at each Outcome circle

  • Choose the best option at each Decision square

  • Finally, recommend the option with the highest expected value

Decision trees as we see are based on expected values, which show the long term average of what will happen if this decision is made many times, therefore they are not ideal for making one off decisions.

It is good to know the risk appetite of the organisation being dealt with, and if they are risk seekers - then we can convey the best possible option that could result, if they are risk avoiders - we can convey the worst possible option that can result.

This would help them better make the decision that suits their risk appetite.

NotesQuizObjective Test