Professional Documents
Culture Documents
Quantitative Techniques
Quantitative Techniques
iii. validating the model – to ensure realistic results of the Types of Probabilities:
experiment;
a. objective probabilities – calculated from either logic or
iv. designing the experiment – involves sampling the actual experience; e.g., the probability that a coin will yield
operation of the system; heads is 50% on any single toss.
v. conducting the simulation and evaluating the results. b. subjective probabilities – estimates of the likelihood of
future events are based on judgment and past experience.
Commonly Used Quantitative Tools:
- e.g., the likelihood that a winner in an amateur singing
a. probability analysis contest will become a successful recording artist.
b. decision tree
d. program evaluation and review technique (PERT) ▪ mutually exclusive – if the two events cannot occur
simultaneously.
e. linear programing
▪ joint probability – both events will occur.
f. queuing
▪ conditional probability – one event will occur given that the
g. learning curves other event has already occurred.
▪ decision-making under conditions of risk – the probability - illustration: White Covered Store sells balut in the city’s
distribution of the possible future states of nature is known. central bus terminal. For this coming weekend, the probability
distribution of the demand for balut is as follows:
Estimated Sales - expected value of perfect information: (EVPI) the difference
Probability Result
in Units between the expected value without perfect information and
750 0.2 150 the result if the best action is taken given perfect information.
900 0.25 225
1,300 0.55 715 - illustration: using the previous example, assume that if Mr.
Expected Value 1,090 B knew the daily sales demand for pansit with certainty, he
would prepare exactly the number of boxes demanded. The
A.2. Payoff Decision Table – presents the outcomes/pay-offs expected value of perfect information is computed as
of specific decisions when certain states of nature (events follows:
which are not controllable by the decision-maker) occur. The
Daily
payoff table is a helpful tool for identifying the best solution CM per Expected
Sales Total CM Probability
given several decision alternatives and future conditions that Box Value
Demand
involve risk. 500 20 10,000 0.2 2,000
600 20 12,000 0.7 8,400
- illustration: Mr. B cooks and sells pansit in a box. Each box
700 20 14,000 0.1 1,400
of pansit is sold for P 50 during regular hours, that is, from 10
Expected CM given Perfect Information 11,800
a.m. to 8 p.m. If every box is sold by 8 p.m., Mr B calls it a day.
However, all unsold boxes by 8 p.m. are sold at half the regular
Expected CM given Perfect Information 11,800
price up to 9 p.m. The variable cost per box is P 30. The
Expected Value of CM using Best Course
contribution margin per box is as follows: (11,500)
of Action without Perfect Information
10 am – 8 pm after 8 pm Expected Value of Perfect Information 300
Selling Price 50 25
Variable Cost (30) (30) Cost of Perfect Information – management may have the
Contribution Margin (Loss) 20 (5) opportunity to acquire additional information that may help in
choosing the best alternative. However, obtaining information
Past experience has shown that the daily sales demand (up requires incurrence of cost.
to 8 p.m.) and their probabilities are as follows:
- in the pansit in a box example, additional information, if
Sales per Day Probability available, would increase the expected value of contribution
500 boxes 0.2 margin by P 300 (from P 11,500 to P 11,800). In this case,
600 boxes 0.7 management would be willing to incur cost of up to P 300 to
700 boxes 0.1 obtain additional information.
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Expected
A. Prepare 500 Boxes: B. Decision Tree – a graphic representation of the decision
Value
(500) 500 x 20 10,000 x 0.2 points, the alternative course of action available to the decision
(600) 500 x 20 10,000 x 0.7 maker, and the possible outcomes from each alternative, as
(700) 500 x 20 10,000 x 0.1 10,000 well as the relative probabilities and the expected values of
each event.
B. Prepare 600 Boxes:
(500) 500 x 20 – 100 x 5 9,500 x 0.2 - node – the intersections in a decision tree.
(600) 600 x 20 12,000 x 0.7
□ – Decision Points (Node) – the points at which the
(700) 600 x 20 12,000 x 0.1 11,500
decision maker must choose some action.
C. Prepare 700 Boxes:
○ – Chance Points (State of Nature Node or Probability
(500) 500 x 20 – 200 x 5 9,000 x 0.2
Node) – the points at which some event related to the
(600) 600 x 20 – 100 x 5 11,500 x 0.7
previous decision will occur.
(700) 700 x 20 14,000 x 0.1 11,250
⎯ – Branches – connectors.
A.3. Expected Value of Perfect Information
- perfection information: the knowledge that a future state of Advantages of Using Decision Tree:
nature will occur with certainty; in this case, it is assumed
that the probability distribution in an accurate representation a. Decision trees facilitate the evaluation of alternatives by
of the relative frequency of future demand and that the giving the decision maker a visual presentation of the expected
decision-maker knows exactly when each possible event will results of each alternative.
occur.
b. Decision trees are useful when sequential decisions are c. A gantt chart can be used to monitor the activities in a
involved. project. It shows which activity should be in progress as of a
certain date and how close it is to completion time.
ii. Determine the events that may result from the chance
points.
v. Evaluate the results and choose the best course of action. D. Program Evaluation and Review Technique (PERT) – a
networking technique used for planning and controlling the
activities in a project. It provides management pertinent
Illustration: Using the pansit in a box example, the decision tree information about a project such as:
is presented as follows:
- expected completion time of the project;
Advantages:
- shortening the total completion time of the whole project d. PERT helps to keep the project on schedule and to provide
can be accomplished by shortening the critical path. feedback to management about the progress of each part of
the project.
Steps in Formulating a Linear Program: (2) Plot the contraints in the problem on a graph.
a. Identify the decision veriables. ▪ locate the terminal points (the x and y intercepts; where one
of the variables is equal to zero)
b. Express the objective and constraint functions in terms of
the decision variables identified in step 1. ▪ join the terminal points by a straight line.
-6B = -144
4G + 2B = 240 4G + 2B = 241
4G = 240 - 48 B = 23.83
---------------------------------------------
4G = 192
4G + 2 (23.83) = 241
G = 48
4G = 241 - 47.66
E (24, 48)
4G = 193.34
(4) Using the objective function, test the four corner points A, B,
C, and D to see which yields the greatest contribution margin: G = 48.335
→ computation of shadown price is an application of F. Queuing Theory (working-line theory) - a study of random
sensitivity analysis. arrivals of a processing or servicing facility of limited capacity;
it allows the decision maker to calculate the:
→ shadow price is the income that would be lost (opportunity
cost) by not adding an additional unit of a scarce resource a. lengths of future waiting lines
(constraint)
b. average time spent in the line awaiting service or
→ in deciding whether to add an additional resource or not, processing
the shadow price must be greater than the accrual cost of
such additional resource. c. additional facilities required
one unit of time. H. Sensitivity Analysis - the study of how the outcome of a
T - average number of work units serviced in decision process changes as one or more of the assumptions
one unit of time (assuming there is no change.