Scribd
Upload
38 views10 pages

Chapter Four

This chapter describes analyzing KFC sales data from 2012-2015 using Bayesian network prediction. The data was formatted in Excel and contained years, months, and quantities. It was input into SPSS Modeler and categorized as nominal data. Quantity was selected as the target for analysis. Three Bayesian network structures - Tan, Markov, and Markov-FS - were used and models were generated. Filter and analysis nodes were used to identify the data contents and generate statistics and graphs of the results, including a graph of predicted quantities over time based on each structure type.

Uploaded by

Shahmi Azmi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
38 views10 pages

Chapter Four

This chapter describes analyzing KFC sales data from 2012-2015 using Bayesian network prediction. The data was formatted in Excel and contained years, months, and quantities. It was input into SPSS Modeler and categorized as nominal data. Quantity was selected as the target for analysis. Three Bayesian network structures - Tan, Markov, and Markov-FS - were used and models were generated. Filter and analysis nodes were used to identify the data contents and generate statistics and graphs of the results, including a graph of predicted quantities over time based on each structure type.

Uploaded by

Shahmi Azmi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 10

CHAPTER FOUR: ANALYSIS

4.1 Introduction

In this chapter, it describes the process of the data collection from KFC Setia City
Mall to generate the outcome of quantity products in year 2012 until 2015. The data
mining method analysis has been used in this study is Prediction by Loan Defaulters
using Bayesian Network.

4.2 Analysis and Result

Figure: KFC Data in Microsoft Excel

Before the data mining been process in SPSS Modeler, the data should be in format
Excel. In this study, we used KFC Data and save in format .scv. The data mining
also easy to process in specific format such as csv, xls, txt and others. KFC Data
have been categorized in three types such as years, months and quantity. In the KFC
Data also have 49 rows including every month in year 2012 until 2015 as shown in
Figure above.
Figure: KFC Data Preview

The first step, the data need to be input var file. Figure shows the has been input into
SPSS Modeler to be process. On right is preview of data mining that has been
generated in table form. If this happen, the data is suitable to be used in the data
mining method and can proceed to the next step.
Figure: Type Node

In figure shown the second step of data mining process. In this part, the data type
will connect with source node. There are 3 types of field representing years, months
and quantity of the product. In measurement, we have decided to select years,
months and quantity as a Nominal data type for KFC data. Quantity has been
selected as a target for data analysis outcome and the others remaining as an input.
Figure: Select Node

Next, in the Select Node there should key in the purpose or outcome that we want in
the end of the data mining process. In this step, we choose “Quantity” because in 4
years back which is in year 2012 until 2015, we want to know how many quantities
of the KFC’s products have been produced. Therefore, the SPSS Modeler will detect
and count only quantities details in each month of years.
Figure: Structure Type

In prediction technique, this part must have involved in data mining process. The
data need to fill out in each of Structure Type as follow the guideline such as Tan,
Markov and Markov-FS. This three structure type also need to connected to the
Select Node to be easy the data link into the structure node.
Figure: Browsing the Model (Nugget)

The three Bayesian Network nodes need to generate the models to produce the
nugget. The nugget model in each structure node which is Tan, Markov and
Markov-FS will be appear separately and it should relate to the Select Node to be
data processing.
Figure: Filter Node

After the it has been connected until Markov-FS, we need to click Filter Node icon
at the bottom to appear in the box process. This step is used to easy the SPSS
Modeler to identify the contents in the data Tan, Markov and Markov-FS to been
show correctly outcome in the end of the data mining process. After it correctly
identify the node, it must be executing the process by clicking the play green button
on top before it proceeds to the next step.
Figure: Outcome (Graph & Statistic)

This is the last step of data mining process as shown in figure . The outcome will be
generating in these two processes such as Analysis and Evaluation. To know the data
is correctly input in the process to the end, it need to execute from beginning data
has been entered until it will generate automatically the statistic of data and the
result.
Figure: Result of KFC Data (Quantity)

In the figure above is shown the graph of quantity. Each line is representative the
structure type such as Tan, Markov and Markov-FS. On the right is shows the
analysis result of the data for output field quantity.

You might also like