MGT555

BUSINESS ANALYTICS

INDIVIDUAL ASSIGNMENT 1

PREPARED BY:
DAYANG TERINI BINTI MASLIM
2019641012
NBO5B

PREPARED FOR:
DR. MUHAMMAD AZMAN BIN IBRAHIM
QUESTION 1

Data collected in the market can be used to describe or inform a certain situation/phenomenon,
to predict or look at trend and to prescribe the course of action. You may search information
from the internet and provide an example of how data being used for descriptive, predictive,
and prescriptive purposes.

(6 marks)

Data collection is the procedure of obtaining, measuring, and analyzing accurate data from
several relevant sources to address issues in research, provide answers to questions, evaluate
results, and predict future trends and possibilities. Almost all research projects include data
collection. The ideal approach to take often depends on the type of data you are collecting but
there are a variety of method to doing so.

To make business decision today, most firms place a strong emphasis on data and for a good
reason. The goal is not just data but if you can not get meaningful insights that results in more
informed action, than facts and data are pointless. Analytics solutions provide a practical
approach to using corporate data. Descriptive, predictive, and prescriptive analytics are the
three most popular types of analytics, and they all work together to give businesses the most
out of their big data. These methods of analysis all offer different points of view.

Business intelligence often begins with descriptive analysis. By providing context to assist
stakeholders in interpreting information, descriptive analysis helps firms understand how they
operate. It collects and organizes historical data through data aggregation and data mining,
creating visualizations such as graphs, charts, reports, and dashboards. Like statistical
modeling, descriptive analysis provides a clear picture of what happened in the past, but it stops
there. It does not offer interpretations or recommendations about what to do going forward.
When running this type of analytics, you will typically start by selecting a KPI as a benchmark
of performance in a specific business sector. After deciding which data sets will be used for
the study and where to get them, you will collect and prepare them. Several techniques,
including patter detection, clustering, summary statistics and regression analysis, will be used
to identify patterns and measure performance. To make data quick and easy to interpret, you
will develop visualization. Descriptive analytics can benefit decision makers from every
department in a company, from finance to operations. For example, the sales team can discover
which customer segments contributed the most money to sales last year, the marketing team
can determine which social media sites generated the highest return on advertising spend during
the previous quarter, the finance team can monitor monthly and annual revenue increase or
downturn and operations team can tract SKU requests over the past year across multiple
geographic locations.

Knowing what happened in the past and understanding why it happened allows you to make
predictions about what might happen in the future. Using historical data, predictive analytics
creates machine learning models that consider important trends and patterns. The model is then
used to predict future events using the latest data. Predictive analytics assesses the likelihood
that certain future events will occur in addition to predicting potential outcomes. It helps firms
in better planning, creating realistic goals and reducing unnecessary risks. Additionally, it
allows teams to more accurately predict future performance based on past performance and all
the elements that currently affects it. What if analysis, where different values are changed to
examine how those changes might affect results, is one of the most useful types of predictive
analytics. Business team are empowered to make better decisions faster when they can perform
rapid iterative analysis to examine options. For example, sales team can discover revenue
prospects of specific customer segments, marketing team can predict how much money an
upcoming campaign is likely to bring in, the finance team may make a more accurate estimate
for the next fiscal year and operations team are better able to anticipate regional demand for
various items at specific times in the coming year.

The next step in predictive analytics is prescriptive analytics. Based on the generated
predictions, this type of analysis directs the team on what to do. Prescriptive analytics predict
the what, when and why of an event or trend. It explains which actions have the best chance of
producing the desired results. It enables teams to address issues, improve performance and
seize profitable opportunities. For example, how sales team can improve sales procedure for
each intended industry, assist in the marketing team’s product selection to highlight the
upcoming quarter, ways finance team can improve risk management and identify ways to
optimize warehousing with the operation team.

Descriptive, predictive, and prescriptive analytics are a potent force when applied together.
Descriptive analytics is an essential technique that helps businesses make sense of vast amounts
of historical data. It helps you monitor performance and trends by tracking KPIs and other
metrics. By combining descriptive analytics with predictive and prescriptive analysis, firm can
gain deeper insights into the causes and likely future outcomes of events, as well as the potential
actions they can take to improve business performance. Like predictive analytics, prescriptive
analytics would not be right hundred percent of the time since they work with estimates.
However, they provide the best way to see into the future and determine the viability of
decisions before making them.
QUESTION 2

The Excel file Credit Risk Data provides information about bank customers who had applied
for loans. The data include the purpose of the loan, checking and savings account balances,
number of months as a customer of the bank, months employed, gender, marital status, age,
housing status and number of years at current residence, job type, and credit-risk classification
by the bank. Answer the following questions by showing your works in spreadsheet of the
given file.

a) Find the measure of central tendency for the checking and savings account balances
among the customers who are classified as high and low credit-risk takers.
(12 marks)

Central Tendency for the Checking Account Balances - High Credit-Risk

Takers

Mean 847.2322275
Standard Error 196.0361426
Median 0
Mode 0
Standard Deviation 2847.589455
Sample Variance 8108765.703
Kurtosis 25.81903539
Skewness 4.91561773
Range 19812
Minimum 0
Maximum 19812
Sum 178766
Count 211

Central Tendency for the Saving Account Balances - High Credit-Risk

Takers

Mean 1487.957346
Standard Error 207.6299382
Median 648
Mode 0
Standard Deviation 3015.999064
Sample Variance 9096250.355
Kurtosis 15.6838512
Skewness 3.826082713
Range 19811
Minimum 0
Maximum 19811
Sum 313959
Count 211
 Central Tendency for the Checking Account Balances - Low Credit-Risk
Takers

Mean 1245.981308
Standard Error 233.2486696
Median 0
Mode 0
Standard Deviation 3412.133872
Sample Variance 11642657.56
Kurtosis 13.40303185
Skewness 3.659976656
Range 19155
Minimum 0
Maximum 19155
Sum 266640
Count 214

Central Tendency for the Saving Account Balances - Low Credit-Risk

Takers

Mean 2132.616822
Standard Error 278.3796859
Median 538.5
Mode 0
Standard Deviation 4072.343723
Sample Variance 16583983.4
Kurtosis 5.728155833
Skewness 2.540846626
Range 19568
Minimum 0
Maximum 19568
Sum 456380
Count 214
 b) Construct a cross-tabulation/pivot table for marital status and housing type and
illustrate the results with a PivotChart. Provide a brief description of it.
(6 marks)

Count of Marital
Status Column Labels
Row Labels Divorced Married Single Grand Total
Other 11 41 52
Own 98 30 164 292
Rent 47 6 28 81
Grand Total 156 36 233 425

From the table, we can see that the number of customers applying for a loan with the
status of single owning their own house is higher which is 164 people, followed by
divorced, 98 and married, 30.
47 divorced customers rent a house followed by single, 28, and married customers, 6.
41 customers with a single status, and 11 divorced customers have other types of
housing.
The highest number of loan applications were from customers with a single status, 233,
followed by divorced, 156, and married, 36.
The total number of loan applications was 425.
 c) Use PivotTables to find the number of loans by different purposes for the different
marital status in the Excel file Credit Risk Data. Illustrate the results on a PivotChart.
(Note: place marital status in column and loan purposes in row)
(6 marks)

Count of Marital
Status Column Labels
Row Labels Divorced Married Single Grand Total
Business 14 4 26 44
Education 13 1 9 23
Furniture 44 5 36 85
Large Appliance 1 1 2 4
New Car 37 8 59 104
Other 2 4 6
Repairs 6 1 5 12
Retraining 1 1 2
Small Appliance 32 13 60 105
Used Car 7 2 31 40
Grand Total 156 36 233 425
 d) Either using slicer or filter, determine the housing type and job type among those bank
customers who are single (show the pivot table and pivot chart). Give a brief description
of it.
(6 marks)

Count of Marital
Status Column Labels
Grand
Row Labels Management Skilled Unemployed Unskilled Total
Other 8 30 1 2 41
Single 8 30 1 2 41
Own 26 98 2 38 164
Single 26 98 2 38 164
Rent 3 17 8 28
Single 3 17 8 28
Grand Total 37 145 3 48 233

From the table, we can see that the highest number of loan applications for single status
is from skilled jobs with a total of 145, followed by unskilled jobs, 48, management
jobs, 37, and unemployed, 3.
164 of the applications own their own houses, 28 applications rent houses, and 41
applications have other types of housing.
The total number of loan applications from single status is 233.
e) Compute the descriptive summary on the savings account balances among the
customers who are married, divorced and single. Compare these groups. What can you
conclude?
(2 marks)

Savings Account Balances

Divorced Married Single

Mean 1884.673077 2000.638889 1735.223176

Standard Error 294.6751482 764.8480409 221.2678787
Median 642.5 529.5 591
Mode 0 0 0
Standard Deviation 3680.491421 4589.088245 3377.507583
Sample Variance 13546017.1 21059730.92 11407557.48
Kurtosis 8.793720842 9.2777815 8.596452666
Skewness 3.014002762 3.14806096 2.964113218
Range 19568 19811 18716
Minimum 0 0 0
Maximum 19568 19811 18716
Sum 294009 72023 404307
Count 156 36 233

In this result, we have 156 observations for divorced, 36 observations for married and
233 observations for single.
Each group, the mean is greater than median, indicates that the distribution is right-
skewed.
The standard deviation of the married group indicates the data are more spread out than
the divorced and single group in which the data are clustered around the mean.
Higher kurtosis in the married group indicates a flatter distribution shape and flatter
tails indicate a greater risk of occasional extreme outcomes.
Therefore, from this result, I can conclude that the single group has more savings
account balances followed by the divorced group. The married group has fewer savings
accounts.