ANALYSIS AND CALCULATIONS

Identify issues and difficulties in the corporate world that are similar
to those described in the case.
• Nowadays it is essential to monitor the performance of business units
on a continual basis.
• A similar approach can be applied to evaluate the performances of
various business verticals
• Such as: hospitals, restaurants and hotels, airlines, insurance
companies, educational institutes.
• Identify various sectors in each business vertical.
• Identify the evaluation criteria for each of these businesses.
TOPSIS

• TOPSIS consists of constructing a global preference relation for a set of

alternatives (also called DMU in DEA) evaluated using various criteria.
• For each of the alternatives and for each of the criteria a mathematical procedure
is carried out to calculate the distance from the best (positive ideal) to the worst
(negative ideal) solution.
• The chosen alternative or the best performing alternative is expected to be close
to the positive ideal solution and far from the negative ideal solution.
• The ideal solution is formed as a composite of the best performance values
exhibited (in the decision matrix) by an alternative for each attribute.
• TOPSIS assumes that there are “m” alternatives and “n” criteria, and we have the
score of each option with respect to each criterion.
Step 1 - TOPSIS
• Let 𝑥𝑖𝑗 be the value for alternative 𝑖 with respect to criterion 𝑗
• Let 𝑋 = (𝑥𝑖𝑗 ) be the (𝑚 × 𝑛) matrix.
• Let T* be the set of benefit criteria (more is better)
• Let T’ be the set of negative criteria (less is better)
Step 2
• Construct a normalized decision matrix to facilitate the comparisons
across criteria. Normalize scores or data are as given by

𝑥𝑖𝑗
𝑟𝑖𝑗 = for 𝑖 = 1, … , 𝑚; 𝑗 = 1, … , 𝑛 (expression 1)
σ 𝑥𝑖𝑗 2
Step 3
• Construct the weighted normalized decision matrix. Assume a set of
weights for each criterion: 𝑤𝑗 for 𝑗 = 1, … , 𝑛
• Multiply each column of the normalized decision matrix by its
associated weight to get 𝑣𝑖𝑗 = 𝑤𝑗 ∗ 𝑟𝑖𝑗

• In the problem under consideration, we consider uniform weights for

all criteria.
• Therefore, this step may be ignored here.
• Next, identify the maximum and minimum values of each criterion.
Step 4
Determine
• positive ideal solution S+(set of maximum values 𝑣𝑗+ with respect to each
criterion)
S+ = {𝑣1+ , 𝑣2+ , … , 𝑣𝑚
+
}
• and negative ideal solution A- (set of minimum values 𝑣𝑗− with respect to
each criterion)
S- = {𝑣1− , 𝑣2− , … , 𝑣𝑚
−}

Where,
𝑣𝑗+ = max 𝑣𝑖𝑗 if 𝑗 is benefit criterion
= {min(𝑣𝑖𝑗 )} if 𝑗 is cost criterion

𝑣𝑗− = {max(𝑣𝑖𝑗 )} if 𝑗 is cost criterion

= {min(𝑣𝑖𝑗 )} if 𝑗 is benefit criterion
Step 5
Calculate the separation measures 𝑆𝑖+ and 𝑆𝑖− for each alternative 𝑖
(DMU)
Where:
• 𝑆𝑖+ is separation measure from the positive ideal solution and
2
𝑆𝑖+ = σ𝑚
𝑗=1 𝑣𝑗+ − 𝑣𝑖𝑗 ∀ 𝑖 = 1,2, … , 𝑛 (Expression 2)
• 𝑆𝑖− is separation measure from the negative ideal solution.
− 2
𝑆𝑖− = σ𝑚
𝑗=1 𝑣𝑖𝑗 − 𝑣𝑗 ∀ 𝑖 = 1,2, … , 𝑛 (Expression 3)
Step 6
• Calculate the closeness coefficient to the ideal solution 𝐶𝑖 for each
alternative 𝑖, where
𝑆𝑖−
𝐶𝑖 = where, 0 ≤ 𝐶𝑖 ≤ 1 (Expression 4)
𝑆𝑖− +𝑆𝑖+

Step 7
• Rank the alternatives or DMUs in the descending order of 𝐶𝑖
• the alternative with maximum value of 𝐶𝑖 is the best performing
alternative.