Decision Science

June 2023 Examination

Ans 1.

Understanding

Bayes’ theorem is a mathematical formula which determines conditional probability. It is

used in the scenario where there is already existing predictions or theories. In this problem
statement, we will use Bayes’ theorem to calculate the number of people associated with bad
mood given they have periodontal disease.

Bayes' theorem mathematical formula is: P(A|B) = P(B|A) * P(A) / P(B)

Here:
P (A|B) is the probability of event A, given event B has occurred.
P (B|A) is the probability of event B given event A has occurred.
P (A) is the probability of event A
P (B) is the possibility of event B

Calculations:
In this problem, we want to find the probability of getting into a bad mood, given that
someone has a periodontal disease., P (bad mood | Periodontal disease)

From the problem statement, following probability can be drawn:

Earlier probability of getting a bad mood = 0.1 (10%)
Probability of having a periodontal ailment given a lousy mood is 0.85 (85%)
Probability of having a periodontal disease given no bad mood is 0.29 (29%)
Below is a tree diagram to represent this information.

First level:
The first level of the tree diagram represents the earlier chance of having a Bad or Healthy
mood. In this case, bad moods are occurring with 10% chance. Therefore, the chance of
having a bad mood is 0.1, and the possibility of having heathy mood is 0.9.

Second level:
The second stage of the tree diagram represents the probability of having periodontal disease
given a bad or healthy mood. The observation suggests the possibility of getting periodontal
disease, given a bad mood, is 0.85. the probability of getting periodontal disease given
healthy mood is 0.29. therefore, the possibility of getting no periodontal disease given a bad
is 0.15, and the opportunity of getting no periodontal disease given no bad mood is 0.71.
Last Level:
Using the tree diagram, we can calculate the probability of getting a bad mood and
periodontal disease, corresponding to outcome 1. From the tree diagram, we can see that the
probability of outcome 1 is:

P (terrible temper and periodontal disease) = P (bad mood) * P(Periodontal disease | bad
mood)
= 0.1 * 0.85
= 0.0.5

P (Periodontal sickness) = P (Periodontal disease | bad mood) * P (bad mood) + P

(Periodontal disease | No bad temper) * P (No bad temper)
= 0.85 * 0.1 + 0.29 * 0.9
= 0.346

Bayes' theorem to calculate the probability of getting a bad mood given periodontal disease:
P (bad mood | Periodontal disorder) = P (Periodontal disease | bad mood) * P(bad mood) /
P(Periodontal disease)
= 0.85 * 0.1 / 0.346
= 0.246

Consequently, the chance of having a bad temper given having a periodontal disease is 0.246
or about 24.6%.

Conclusion
Using Bayes' theorem and a tree diagram, we've determined that the probability of a person
having a bad temper, given periodontal disorder, is about 24.6%. This result shows a certain
correlation between periodontal disease and bad moods, with a significantly probability of the
bad mood among the person with the Periodontal disorder. So, it becomes clearly important
from this calculation of the oral care as it can be important parameter for the mood
regulation.
Ans 2.

Understanding

Linear regression analyses the data of dependent and independent variable. It visualises the
relationship between two variables. Linear regression is a data analysis technique that
predicts the value of unknown data by using another related and known data value.

Procedure:
1. Input the records in two columns, with the independent variable (number of posts per day)
in one column and the dependent variable (no of followers) in another.

2. Select the statistics tab and select the data evaluation.

3. Select Regression from the listing of analysis equipment and select adequate.

4. Input the data in the Regression conversation box.

5. Run the regression analysis.

6.Output will be displayed.

Input
Output

Interpretation of Excel Tables:

Regression statistics:

Multiple R: The value of 0.0466 shows that the correlation between the variables is weak.

R square: This low value shows that the model may need to fit higher data.

Adjusted R -square: negative value indicates that the model does not fit the records well.

Standard error: It suggests the average distance between the actual values and the anticipated
values.

The ANOVA table gives information on the significance of the regression version. Slope
Coefficient and Intercept value for the linear regression fit line can be obtained from this
table.
The intercept of 377.21 when the x variable is zero. The coefficient of the number of posts
according to day is 1.74.

So, using these, the regression equation is constructed as: y = 377.21+ 1.74 x

Conclusion

The low R-square, Adjusted R-square values, and low F-statistic implies that the version does
not suit the information nicely. the p-value for this coefficient is insignificant, indicating that
it is not statistically different from zero. Also, the number of the observation that we feed for
analysis are not so many in number so the model not able to train so perfectly. Also, the
nature of the data maybe not so suitable for getting the linearity among them. Maybe this data
set fit more suitably to other regression models like SVM, Decision Tree etc.

Ans 3a.

In Interval of the replacements of 1000 light bulbs mounted in a new factory needs to be
fined out in this problem. Bulb has an average lifetime of 120 days and a standard deviation
of 20 days and that we cannot allow extra than 10% of the bulbs to expire earlier than
replacement.

Now assume that X is the bulb's lifetime in days, and then X is normally distributed with a
mean of 120 days and a standard deviation of 20 days.

We want to find the value x such

P (X <= x) = 0.10.
In the first step value of the X is standardize

Z = (X - μ) / σ

Where μ is the mean and σ is the standard deviation.

Z = (x - 120) / 20

Using the standard regular distribution table, we find that the z-score corresponding to P (Z
<= z) = 0.10 is -1.28.

Substituting into the formula, we have

-1.28 = (x - 120) / 20

x = -1.28 * 20 + 120 = 94.4

Therefore, the interval between replacements should be 94 days or less to ensure that no more
than 10% of the bulbs expire before replacement.

Conclusion

From the above calculation, it can be concluded that interval between replacements for 1000
light bulbs with a mean lifetime of 120 days and a preferred deviation of 20 days must be
94.4 days or less to ensure that no extra than 10% of the bulbs expire before replacement.
Normal distribution statistics are used for the statistical purpose for getting the value. Also
replacement in the given time frame is important for the productivity and the safety of the
work force.
Ans 3b.

To get the average age of migrants for both the categories of gender, we need to find the
figure by dividing the integrated number of the mid-range of every age group with the
number of migrants in that group to the total population of the given gender in the table.
From the above solved table, we the

Average age of male migrants = (Sum of f*x) / (Total number of Male migrants)
= 4,613,678,689 / 1, 455, 99803
= 31.687

Average age of male migrants is approx. 31.69 years.

Average age of female migrants = (Sum of f*x) / (Total number of Female migrants)
= 11,316,123,244 / 30, 859, 1233
= 36.670

Average age of female migrants is approximately 36.67 years.

Interpretation:

From the above calculation, we got the average age of male migrants to be 31.687 years, and
the average age of girl migrants to be 36.670 years. Now the difference within the typical age
can be due to various reason:

In general, the male migration is due to the fact of the early opportunities for the better job
and hence they try for the migration to get settle for the career opportunities out there. On the
other hand, female migration is generally linked with the family unification role, hence it
happened in the later age of the life after the male counterpart gets settled out there.

But if we look into the total number of the migration, female numbers come out to be on the
higher side. Female numbers are generally higher for the procurement of the higher education
from the abroad and hence t give push to the total female migrants out there. Also, from the
less contribution from male side for the family unification taunts its total number in the
migration stats.