Quantitative Methods in SOM
Data Envelopment Analysis (DEA)
• Technique
– based on linear algebra, lpp concepts
– handle complex relationships between multiple inputs
and multiple outputs
– discover inefficient opns / units
– analyse-benchmark-compare-evaluate performances of
several service units like banks, hospitals, schools etc
DEA Process
• Measure performance,efficiency,productivity in the form of ratio of Output and Input to some processes
• Evaluating activities of service org
• Productivity can be measure by two methods.
– Partial productivity measures
– Total factor productivity measures √
• Partial productivity measures does not consider all output and input factors, whereas, total factor
productivity measure can take into account of all the outputs and inputs. Therefore the mistake of
imputing gains to one output that are attributable to another output in partial productivity measures
can be avoided using total factor productivity measures.
• To consider all inputs and outputs, a tool is required which can manage following challenges.
– How to get a single ratio considering multiple outputs and multiple inputs?
– How to give importance to one attribute over other: Weights!
– How to handle large number of variables and constraints?
• To understand DEA methodology we will first consider single input and single output case.
Single input and single output
• Consider an example of a chain of coffee shops ’ located in eight locations from A to H
• Owner wants to evaluate the efficiency of the shops in one city.
• He considers
– cups of coffee sold (per day) as output
– number of employees in the store as inputs.
• owner wants to know
– which store is efficient and which store is inefficient
– Benchmark best store so that he can suggest improvements for inefficient stores by comparing
Comparison
Result
• plot input and output variables on a graph, eight stores can be located on a graph as shown
• line segment touching all the store points and origin represents the slopes all the store
points have.
• store B has max efficiency
• If we extend the line segment from origin to store B, it will envelope all other points
because it has maximum slope and efficiency.
• This line segment is called efficient frontier
Efficient Frontier
• line connecting the point (0,0) and the most efficient point is the efficient frontier
called ‘frontier line’. Efficient Frontier envelopes other data points, that’s why it is
called Data Envelopment Analysis (DEA). Here, in the example point B is the efficient
frontier.
• frontier line displays the performance of the best store in the comparison.
• efficiency of other stores can be measured by the deviation of the points from the
frontier line. Efficiency of other stores is measured relative to the efficient frontier.
• Efficient frontier serves as Benchmark.
Relative efficiency of all stores with respect to most efficient store
• The relative efficiency of stores is measured by
taking the ratio of efficiency of each store and the
efficiency of most efficient store as shown below.
• 0≤ 𝑁os 𝑜𝑓 𝑐𝑢𝑝𝑠 𝑜𝑓 𝑐𝑜𝑓e𝑒 𝑠𝑜𝑙𝑑 𝑝𝑒𝑟 𝑛os 𝑜𝑓 𝑒𝑚𝑝𝑙𝑜𝑦𝑒𝑠 𝑓𝑜𝑟 𝑒𝑎𝑐ℎ 𝑠𝑡𝑜𝑟𝑒 ≤1
𝑁os 𝑜𝑓 𝑐𝑢𝑝𝑠 𝑜𝑓 𝑐𝑜𝑓𝑒e 𝑠𝑜𝑙𝑑 𝑝𝑒𝑟 𝑛os 𝑜𝑓 𝑒𝑚𝑝𝑙𝑜𝑦𝑒e𝑠 𝑓𝑜𝑟 𝑚𝑜𝑠𝑡 𝑒𝑓𝑖𝑐𝑖𝑒𝑛𝑡 𝑠𝑡𝑜𝑟𝑒
The relative efficiency score are free from units of measure (u nits invariance) that’s why it can
be used to compare different stores regardless of units. The comparison moves from Ratios to
the ratio of ratios. Let us consider the example of improving the efficiency of store A. This can be
done by increasing the number of coffees sold or by reducing the number of employees
employed in the store A.
Multiple inputs and one output
• Now owner likes to see the impact of one more input that is floor
space along with number of employees on the performance of each
store.
• Using the concept of constant return to scale, convert the values of
input to the amount of input required to obtain a unit output
• DEA will be used to evaluate stores which use less input resource to get
one unit of output.
Multiple inputs and one output
Multiple inputs and one output
• For multiple inputs case we will determine the efficient
frontier by selecting the points or stores, for which one input
cannot be improved without worsening the other input.
• efficient frontier is the line connecting EDC
• points E, D and C cannot improve one of its input values
without worsening the other input. This line envelops all the
other units. This set of points is called as the Production
Possibility Set.
• The observed points are assumed to provide (empirical)
evidence that production is possible at the rates specified by
the coordinates of any point in this region.
Result & Discussion
• Let us consider a unit and see how we can measure the amount of improvement needed
for inefficient store or a unit to make to reach the efficient frontier.
• Consider the point A. Connect the point A to the origin. It would cross the efficient frontier
at a point and let us call it P
• efficiency of A can be computed by the lengths OP and OA. Efficiency of A = 0.8751. This
value can be improved by dropping a line along the X-axis and the Y-axis from the point A
to the efficient frontier. Thus the unit can strive to achieve efficiency equal to the ones on
the efficient frontier.
Single input, multiple outputs
• Let us consider the case when there is one input and two outputs. A
retail giant has 7 stores in a metro city. The owner of retail giant wants
to compare the performance of the retail stores. The owner would like
to consider input resources to be the number of salesman (x) and two
outputs; number of footfalls or customers visiting those stores (y1) and
revenue earned (y2) by 7 stores. The values observed are presented as
shown
• Similar to the previous case, the efficiency of a
particular store can be obtained by connecting a
line from the origin to the point of the store,
crossing the efficiency frontier. Let’s calculate the
efficiency of D as presented .
• Efficiency of D = distance (O,D) / distance (O,P) =
0.75.
Discussion
• This ratio 0.75 depicts the proportion of the output that ‘P’
shows was possible of achievement and the proportion of
inefficiency present in both outputs by D.
• • The store D can be made efficient (as efficient as the stores
in the efficient frontier) by maintaining the same ratio of
inputs to output by moving along the lone (O, P).
• • The inefficiency which can be eliminated without changing
the proportions is referred to as “Technical Inefficiency”.
• • If we have to alter the proportion of outputs, it is referred
as “Mix Inefficiency” similar to the case of unit A in the
example.
Multiple inputs, multiple outputs
• Now, consider the case when there is more than one input affecting more than one output. The
efficiency in such case cannot be determined like in the previous case of single input and single output.
• To evaluate the efficiency of the shops in one city.
• items sold as outputs in his store are in two categories: ‘Beverages – Coffee, Tea and other drinks’ and
‘Snacks – Sandwiches, Biscuits and so on’.
• inputs leading to the sale of these items are: Available floor space in the store and the number of
employees.
Multiple inputs, multiple outputs
• To find the efficiency of each of the units and to
determine the relative efficiency, we need to give weights
to each input and to each output. These weights are the
coefficients for each input and output variable. The
coefficients pertaining to output variables measure the
relative decrease in efficiency with each unit reduction of
output variable. The coefficients pertaining to input
variables measure the relative increase in efficiency with
each unit reduction of input variable. These weights can
be either provided by the user or can be determined
optimally by using some optimization technique like DEA.
Multiple inputs, multiple outputs
CCR Model(1978)
Example
• ‘Super labs’ are in Research and Development of electronic goods. The founders, in the year 2001 decided to
set-up their own company to produce consumer electronic goods that use the state-of-the art technologies
that have been developed by them. The company has set up business in North America and quickly moved to
Asia, Europe, Africa and South America. The main products of the company are its impressive range of Smart
phones, Tablet PCs and Laptops.
• At the beginning of the year 2011, the CEO- global operations of the company quickly pulled reports of Inputs
and Outputs that are being used and produced at various regions in which company operates.
• He arrived at the following figures:
• The data tells him how much each of the regions consume for the two main activities,
namely Product development and Marketing, and how many units of its products are
being sold. However, he is unable to figure out, which are the best performing units and
how much of increase or decrease the Regional directors are to be advised to make on
investments.
• To analyze the performance of the different units and then make a conclusion.
• inputs and outputs to the various ‘Decision Making Units’ are the same units ($ spent of
product development, marketing and number of units of each product sold).
Optimized Solution
Result
• Thus on comparing the inputs and outputs of Europe with that of N and S
America, Europe to reduce Product development and Marketing costs by
3.302 and 5.947 (‘000s of dollar value) and increase sales of Tablets and
Laptops by 1.256 and 2.71 units respectively to reach the values of the
efficient units. The sale of Smart phones is however to the level of the
efficient units N and S America. Thus the changes to be made to the other
inefficient DMUs can be analyzed.
Simulation
• Simulation is a powerful technique for solving a wide
variety of problems. Simulation is an imitation of the
operation of a real world processor system over time.
Simulation involves the generation of an artificial history
of a system and the observation of that artificial history
to draw inferences concerning of the operating
characteristics of the real system. Simulation is done
either manually or using computers. Simulation is
basically an experimental technique. It is a fast and
relatively inexpensive method of doing an experiment on
the computer.
Examples
• Simulating aircraft delay absorption
• Runway schedule determination by simulation optimization
• Modeling ship arrivals in ports
• Modeling and simulation of telephonic call centre
• Baggage screening at airports
• Telecommunication billing system
• Modeling front office and patient care in ambulatory health care
practices
• Project management
• Scheduling of police patrols
• Hazardous waste handling
Questions
