Decision Tree
Decision Tree
Decision Tree
-$0.06m
$2.4m
$1.6m - $1m = $0.6m
$1.6m
-$0.8m
Decision trees - explanations
The square These are the estimated
highlights that a expenses if this course of
decision has to action is chosen.
be made
square =
decision nodes The financial result/revenue
if the course of action
succeeds or fails.
British Nelson is an oil-producing company. There are two parts to the business, searching for new oil
fields and oil extraction. The company currently has low retained profits as they have recently paid a
huge fine to the USA government.
1. Read the given information and based on it construct a complex decision tree with 4 decision nodes.
2. Based on your calculations of EMV and net gain, answer the question above.
Case study: British Nelson
Where should they look for a new oil field?
1. Read the given information and based on it construct a complex decision tree with 4 decision nodes.
2. Based on your calculations of EMV and net gain, answer the question above.
1. Make sure you have enough space around the chance nodes and decision nodes as there is a lot of
information you need to squeeze in there.
2. Make sure that your decision tree is concise and it could be easily understood. Think of the most logical
space to put your calculations (EMV and net gain).
3. When you find the net gain for selling and developing the fields, you need to start eliminating the
options (whether developing or selling is best). Then go back to the success or fail of finding an oil field in
the particular region where you should again find the net gain which is the one you will consider at the
end.
Case study: British Nelson C he ckin
g
1. Check if you have constructed correctly the decision tree by taking a look at the answer key.
1. Based on your personal experience of constructing a decision tree, elicit possible advantages
and disadvantages of using a decision tree.
Take a look at the questions on the board and answer them within your group:
● Reflect on your last exam and set goals for improving your work.
● Construct a decision tree.
● Explain the requirements for the IA in terms of structure.