decision analysis
Transcript of decision analysis
Table of content
Case Study..............................2
1.0 Definition of Decision Tree.........3
2.0 The certainty and uncertainty in
decision-making environment.............5
3.0 The principal steps of decision
analysis................................7
4.0 Decision Tree step-by-step..........8
5.0 Benefits of using decision analysis 14
6.0 Summary............................17
7.0 References.........................18
8.0 Coursework.........................19
Page 1 of 32
Case Study
Patrick Oil Company has leased the drilling rights on a large
parcel of land in Tawke that may not contain an oil reserve. A
competitor has offered to lease the land for $200,000 cash in
return for drilling rights and all rights to any oil that
might be found. The offer will expire in three days. If
Patrick does not take the deal, it will be faced with the
decision of whether to drill for oil on its own. Drilling
costs are projected to be $400,000. The company feels that
there are four possible outcomes from drilling:
1. dry hole (no oil or natural gas)
2. natural gas
3. natural gas and some oil
4. oil only
Page 2 of 32
If drilling yields a dry hole, the land will be basically
worthless, because it is located in the badlands of Tawke. If
natural gas is discovered, Patrick will recover only its
drilling costs. If natural gas and some oil are discovered,
revenue is projected to be $800,000. Finally, if only oil is
discovered, revenues will be $1600, 000.
Assignment:
(a) Describe the certainty and uncertainty in decision-
making environments.
(b) Structure the decision problem as a decision tree.
(c) Solve for Patrick Oil Company’ optimal decision
strategy.
1.0 Definition of Decision Tree
Decision Trees are excellent tools for helping you to choose
between several courses of action. They provide a highly
Page 3 of 32
effective structure within which you can lay out options and
investigate the possible outcomes of choosing those options.
They also help you to form a balanced picture of the risks and
rewards associated with each possible course of action.
In another words, decision tree is a schematic tree-shaped
diagram used to determine a course of action or show a
statistical probability. Each branch of the decision tree
represents a possible decision or occurrence. The tree
structure shows how one choice leads to the next, and the use
of branches indicates that each option is mutually exclusive.
Decision tree learning uses a decision tree as a predictive
model which maps observations about an item to conclusions
about the item's target value. It is one of the predictive
modeling approaches used in statistics, data
mining and machine learning. More descriptive names for such
tree models are classification trees or regression trees. In
these tree structures, leaves represent class labels and
branches represent conjunctions of features that lead to those
class labels.
Page 4 of 32
In decision analysis, a decision tree can be used to visually
and explicitly represent decisions and decision making.
In data mining, a decision tree describes data but not
decisions; rather the resulting classification tree can be an
input for decision making.
In my opinion, decision trees provide an effective method of
Decision Making because they are clearly lay out the problem
so that all options can be challenged. Besides that, decision
trees allow us to analyze fully the possible consequences of a
decision. It also provides a framework to quantify the values
of outcomes and the probabilities of achieving them. Last but
not lease, decision tress help us to make the best decisions
on the basis of existing information and best guesses.
As with all Decision Making methods, decision tree analysis
should be used in conjunction with common sense. This is why
Page 5 of 32
decision trees are just one important part of your Decision
Making tool kit.
2.0 The certainty and uncertainty in
decision-making environment
Page 6 of 32
The decisions are taken in different types of environment. The
type of environment also influences the way the decision is
made.
There are two types of environment in which decisions are
made.
1. Certainty:
In this type of decision making environment, there is only one
type of event that can take place. It is very difficult to
find complete certainty in most of the business decisions.
However, in many routine type of decisions, almost complete
certainty can be noticed. These decisions, generally, are of
very little significance to the success of business.
Condition under certainty is which the decision maker has full
and needed information to make a decision. Decision is made
under the condition of certainty. The manager knows exactly
what the outcome will be, as he or she has enough clarity
about the situation and knows the resources, time available
for decision-making, the nature of the problem itself,
Page 7 of 32
possible alternatives to resolve the problem, and undoubtedly
clarify or certain with the result of alternatives.
In most situations, the solutions are already available from
the past experiences or incidents and are appropriate for the
problem at hand. The decision to restock food supply, for
example, when the goods in stock fall below a determined level
is a decision-making under circumstances of certainty.
2. Uncertainty:
In the environment of uncertainty, more than one type of event
can take place and the decision maker is completely in dark
regarding the event that is likely to take place. The decision
maker is not in a position, even to assign the probabilities
of happening of the events.
In a situation of uncertainty, on the other hand, people have
only a meager database, they do not know whether or not the
data are reliable, and they are very unsure about whether or
not the situation may change.
Page 8 of 32
Moreover, they cannot evaluate the interactions of the
different variables. For example, a corporation that decides
to expand its Operation to an unfamiliar country may know
little about the country’s culture, laws, economic
environment, and politics. The political situation may be
volatile that even experts cannot predict a possible change in
government.
Such situations generally arise in cases where happening of
the event is determined by external factors. For example,
demand for the product, moves of competitors and so on are the
factors that involve uncertainty.
Page 9 of 32
3.0 The principal steps of decision
analysis
To solve a decision tree problem, the following steps are
used.
Step 1: Grow a decision tree.
Arrange the decisions and events in the order in which they
will occur. This is often difficult with complex decision
problem with complex decision problems, but unless the
decision tree accurately represents the situation, the
decision made may not be the best decision.
Step 2: Assign the possibilities.
Make the necessary probability assessments and show them on
the event branches. These probabilities can be determined
using classical assessment, relative frequency of occurrence,
or subjective techniques. (Remember that probabilities are
associated with the uncertain events and not with the decision
alternatives.)
Page 10 of 32
Step 3: Assign cash flow
Assign cash flow by showing costs and payoffs on the branches
where they occur. Accumulate these cash flows and determine
the end value for each branch of the decision tree.
Step 4: Fold back the decision tree
At each decision fork, select the decision that maximizes
expected payoffs or minimize expected costs.
4.0 Decision Tree step-by-step
Step 1: Grow the decision tree.
The decision tree is developed by organizing the decisions and
events in chronological order. In this example, the initial
decision to be made is whether to accept the lease. The tree
is started as:
Page 11 of 32
If the land is leased, no further decisions are required.
However, if the land is not leased, Patrick faces the decision
of whether to drill on the property. The tree then grows to:
Now, if Patrick decides to drill, there are four possible
events that could occur. These are shown on the decision tree.
Page 12 of 32
Patrick Oil Company
lease out the land
do not lease out the land
Patrick Oil Company
lease out the land
do not lease out the land
drill
not drill
When finished, the decision tree should show all the decisions
and events.
Page 13 of 32
Patrtick oil Company
lease out the land
do not lease out the land
do not drill
drill
dry hole
natural gas
natural gas and some
oil
oil only
Step 2: Assign probabilities to the event outcomes on the
tree.
In this example, the only event deals with the production
result if Patrick decides to drill. The company has
subjectively assessed of probability of each of the four
possible outcomes as follow:
Outcomes Probabil
ity
Dry hole 0.2
Natural gas only 0.4
Natural gas and
some oil
0.3
Oil only 0.1
Page 14 of 32
Step 3: Assign the cash flows to the tree.
At each branch of the tree at which revenue or a cost occurs,
show the dollar value. These revenues and costs are then
totaled across the tree and the end values for each branch are
determined. These cash flows are placed on the tree as
follows.
Page 15 of 32
Patrtick oil Company
lease out the land
do not lease out the land
do not drill
drill
0.2 dry hole
0.4 natural gas
0.3 natural gas and some oil
0.1 oil only
END VALUE
Page 16 of 32
Patrtick oil
Company
$200,000lease
out the land
$200,000
do not lease
out the land
do not drill
$0
-$400,000 drill
$00.2 dry hole
-$400,000
$400,0000.4
natural gas
$0$800,00
00.3
natural gas and some oil
$400,000
$1600,000
0.1 oil only
$1200.000
Step 4: Fold back the decision tree and compute the expected
values for each decision.
We need to compute the expected value for each decision
alternative. This is done starting from the right side of the
tree and working back to the left. We first determine the
expected value for the Drill branch as follows:
E [Drill] = $400,000(0.20) + $0(0.40) + $400,000(0.30) +
$1,200,000(0.10)
= $160,000
As we fold back the tree, we block all decision alternatives
that do not have the highest expected value. This is shown in
the decision tree as follows.
END VALUE
Page 17 of 32
Patrtick oil
Company
$200,000lease
out the land
$200,000
do not lease
out the land
do not drill
$0
-$400,000 drill
$00.2 dry hole
-$400,000
$400,0000.4
natural gas
$0$800,00
00.3
natural gas and some oil
$400,000
$1600,000
0.1 oil only
$1200.000
Note that we always select decisions with the highest
expectation payoff. In this example, the best decision is
to lease the land and accept $200,000 payment because
it exceed $160,000 expected value of the not non-
lease option.
Page 18 of 32
Patrtick oil
Company
$200,000lease
out the land
$200,000
do not lease
out the land
do not drill
$0
-$400,000 drill
$00.2 dry hole
-$400,000
$400,0000.4
natural gas
$0$800,00
00.3
natural gas and some oil
$400,000
$1600,000
0.1 oil only
$1200.000
There are couples of benefits of using a decision tree. First
of foremost, it is very easy to understand and interpret for
any reader. To better understand how decision trees work, it
is best to consider some examples. The decision tree below
provides its students are on a company trying to decide
whether or not to invest in an online training program.
The main decision to be made, which starts in the first box,
is whether or not to spend money on a new training program.
From that box, the two options are yes or no, so one branch is
Page 20 of 32
drawn from the first box with a "yes" on it and another with a
"no" written on it. Since the outcome from each is still
unknown, circles are drawn at the end of each branch
From the yes branch, there are then two options for different
training programs, system 1 and system 2. So, from the circle
at the end of the yes branch, two more branches are drawn: one
for system one and the other for system two. From there,
branches are drawn out of for how the system will be paid for,
either in full or monthly installments.
On the "no" side of the tree, branches are drawn to continue
using the existing training method or hire live training
experts. Since the existing plan is already in place, no more
branches are drawn out from that branch. From the hiring live
trainers branch, two more branches are drawn for those to be
either field technicians or sales engineers.
Once the tree is complete, the costs and probabilities are
assigned to each branch in order to calculate and evaluate
each option.
Page 21 of 32
By building the decision tree, small details that may have
been missed are taken into consideration. For the examples,
since there are a lot of calculations involved in creating
decision trees, many businesses use dedicated decision tree
software to help them with the process. Decision tree software
helps businesses draw out their trees, assigns value and
probabilities to each branch and analyzes each option.
An addition to that, decision trees require relatively little
effort from users for data preparation. To overcome scale
differences between parameters - for example if we have a
dataset which measures revenue in millions and loan age in
years, say; this will require some form of normalization and
scaling before we can fit a regression model and interpret the
coefficients. Such variable transformations are not required
with decision trees because the tree structure will remain the
same with or without the transformation. Another feature which
saves data prep time: missing values will not prevent
splitting the data for building trees. This article describes
how decision trees are built. Decision trees are also not
Page 22 of 32
sensitive to outliers since the splitting happens based on
proportion of samples within the split ranges and not on
absolute values.
Decision trees help in brainstorming outcomes. Decision trees
help you think of all possible outcomes for an upcoming
choice. The consequences of each outcome must be fully
explored, so no details are missed. Taking the time to
brainstorm prevents overreactions to any one variable. The
graphical depiction of various alternatives makes them easier
to compare with each other. The decision tree also adds
transparency to the process. An independent party can see
exactly how a particular decision was made.
Another benefit of decision tree is displaying the sequencing
and interrelations of tasks and events. The schematic display
of starting events, secondary and terminating events allow for
insights into input / output relationship and start/stop
phasing as branching is extending into the future. Priorities
can be established from the difficulties, complexities, and
time requirements suggested by each path.
Page 23 of 32
6.0 Summary
Through this assignment, I am now know essentially all of the
important points related to learning decision trees as well as
many points seminal to learning in general.
First and foremost, I know that learning a classifier can be
useful in decision making. Next, I know what a decision tree
is. Besides that, I have learn that how to write a basic
decision tree analysis step by step. Last but not lease, I
Page 24 of 32
know how to make decision accurately by constructing a
decision tree when facing the conflicts.
In conclusion, decision tree analysis really guides me a lot
in term of making my decision. It is useful for me for the
future when I own a company.
Of course, my point of views is just the tip of the iceberg;
there are many other areas to investigate such as some fairly
well resolved, while others remain topics of very active
research, which will continue to lead to interesting, and
financially rewarding, results.
Page 25 of 32
7.0 References
http://www.investopedia.com/terms/d/decision-tree.asp
http://www.mindtools.com/dectree.html
http://en.wikipedia.org/wiki/Decision_tree_learning
http://www.businessnewsdaily.com/6147-decision-tree.html
http://www.simafore.com/blog/bid/62333/4-key-advantages-of-
using-decision-trees-for-predictive-analytics
http://smallbusiness.chron.com/advantages-decision-trees-
75226.html
http://decisiontreemodel.blogspot.com/
http://www.clt.astate.edu/asyamil/groebner6ed/ppt/chap18.pdf
http://webdocs.cs.ualberta.ca/~aixplore/learning/
DecisionTrees/InterArticle/9-DecisionTree.html
http://www.yourarticlelibrary.com/decision-making/decisions-
making-environments-certainty-uncertainty-and-risk/10269/
Page 26 of 32
http://www.studymode.com/essays/Decision-Making-Under-
Certainty-Uncertainty-And-153805.html
8.0 Coursework
NAME: CHOCK MEI YAN
NRIC: 930330-10-5616
H/P NUMBER: +6016-9827381
STUDENT ID: 200106
1. Decision Management Systems are differing from
traditional Decision Support System in five ways. Explain
carefully.
Decision Support Systems provide information that
describes the situation and perhaps historical
trends so that humans can decide what to do and
which actions to take. Decision Management Systems
Page 27 of 32
automate or recommend the actions that should be
taken based on the information that is available at
the time the decision is being made.
The policies, regulations, and best practices that
determine the best action are embedded, at least in
part, in a Decision- Management System where a
Decision Support System requires the user to
remember them or look them up separately.
The information and insight presented in a Decision
Support System is typically backward looking, and
Decision Support Systems are generally reactive—
helping human decision-makers react to a new or
changed situation by presenting information that
might help them make a decision. In contrast,
Decision Management Systems use information to make
predictions and aim to be proactive.
Learning is something that happens outside a
Decision Support System and inside a Decision
Management System. Users of Decision Support Systems
are expected to learn what works and what does not
Page 28 of 32
work and to apply what they learn to future
decisions. Decision Management Systems have
experimentation or test-and-Learn infrastructure
built in so that the system itself learns what works
and what does not.
Decision Management Systems are integrated into an
organization's runtime environment. They make
decisions for applications and services in the
organization's enterprise application architecture.
In contrast, Decision Support Systems are often
desktop or interactive applications that execute
outside the core application portfolio.
Page 29 of 32
2. A Business-centric rules management environment has a
number of characteristics. Explain carefully.
Each business expert sees only the business rules for
which he is responsible:
Multiple business experts might use the rule
management environment, and they should be able to
read those rules they have read permissions for and
change those rules they have read/write permissions
for. They should not have to navigate through a lot
of other rules or rule repository structures to find
them.
These business rules are presented in context
Business users are making these changes in a business
context—for instance, they are responding to new
regulations or trying to improve the performance of
the decision service. This context should be
reflected in the rule management environment so that
Page 30 of 32
changing the rules feels like just part of running
the business.
The business rules editing environment allows only
those changes that make sense:
Some rules can only be changed in certain ways
because of the underlying data—only certain values
can be set as a consequence of a rule, for
instance. Conditions and consequences that make no
business or technical sense should not be allowed;
if the overall decision service constrains the
specific rules in question to behave within a
range of allowed behaviour, then this should also
be enforced.
The business expert can rapidly see the impact of her
proposed changes: The rule management environment
should be linked to the impact analysis tools and
techniques described later.
No unnecessary technical information is presented
Page 31 of 32