1

Week 2 Assignment

Week 2 Assignment

Sri Ram Kommana

Business Intelligence (ITS-531-A04)

Dr. Jason Hutcheson

University of the Cumberlands

May 19th, 2024

Week 2 Assignment

Chapter 3

Discussion Question

Question 1

Data is the center of all analytics by offering the critical information which is needed

by analysts to use for their decision-making processes. Before any analytics can be

performed, data is required to establish a foundation from which patterns, trends, and

correlation can be identified, which are fundamental things need for judgment making in

business (Klee et al. , 2021). Analitical data capacities allow the companies to convert the

meaningless data to the business intelligence with the aim to make critical business decisions

and optimise the operational effectiveness. Thus, the part played by data in analytics is

irreplaceable, which proves the fact that it is the data which helps to disclose informative

insights and to create the business value.

To think of analytics without data for any analytical process would be ridiculous for

data is the bare foundation that it stands on. The analysis of data includes the application of

systematic computational process that uncovers patterns and insights (Klee et al. , 2021).

Without data, we cannot have any data to analyze. Since there is no data, there will be

nothing to do. Synergy between data and analytics is what allows organizations to leverage

the information power and pull it towards their competitive advantage making a case out for

data as an essential element of any analytical framework.

Question 2

Developing a fresh and extensive concept of business analytics, the stuff for the

analytics cycle represents data, technology, and analytical expertise. Data comes to be a basic

source of input, which are further processed to be utilised as raw material for analysis (Klee
3

Week 2 Assignment

et al. , 2021). Technologies, which are analysis and numbers processing tools, make it

possible to convert the initial raw data into decision-making information. Quantitative

thinking, with a close knowledge of statistics, utilization of systems and data analysis is

required for drawing the business intelligence.

Analytics continuum product is warehouse characterized by numbers, pattern

detection and action. Insights, which in this case are to do with the profitable and strategically

important patterns and trends derived from the data, help in decision-making (Klee et al. ,

2021). This course of actions, which are driven by information collectd from analytical

monitoring tools, leads to better indicators of performance and competitive advantage. As a

whole, the goal is to continue building through improvement in procedures, strategic

restructuring and by fostering growth and sustainability efforts through innovation.

Question 3

There has the inherent capacity in that entity to access both internal and external

information resources for the analytics needs in business. Internal avenues use T-Channel,

XYZ reports, and CRM systems that provide the structured data for customer relationships,

inventory, as well as sales (Klee, Van Guler, and Durbin, 2021). Principally, it means the data

compiled by firms do their research and consumers’ ratings found on social media or where

public reviews are stored as well as semi-structured andunstructed data collected by agencies

and institutions which include but not limited tomarkets reports and economic indexes.

The churning in data can be achieved via different sectors such as business

intelligence tools, machine learning algorithms, and programmatic advertising. Data

structures are usually ordered in a form that simplifies the search process because they are

precisely defined, and majority of them are dependent on the relational database (Klee &

Moon, 2021). The disorderly data are referred to the collection of unstructured (for instance,
4

Week 2 Assignment

text, images, and videos) information that is not predictable, and this data requires more

complicate processing methods to acquire and store it. On the one hand, the JSON or XML

documents which also known as semi-structured data, have some features signifying the

structured data and some that signifies the unstructured data too. With that, a huge global

business analytics system comes to the market so that people can base their decisions on

presented data by different sources.

Question 4

For companies to have data which is analytic-ready for their analysis, the key metrics

that they are supposed to prioritize includes accuracy, completeness, consistency, timeliness,

and relevance. Dependability describes the hits of the data to be correct and remain the scenes

which is reflected in the real world of this scenario (Klee et al. , 2021). Complete data, all

data that are required and there is no missing value, is a priority for good quality data.

Consideration of the mechanisms that the data is uniformly clean within different datasets

and systems being checked, in order to stop the inconsistency that may result in incorrect

analysis.

Gersemness serves to hold the data as refreshed in which they can be used for current

analysis and decision-making (Klee et al. 2021). Generality emphasizes the fact that data is

tailored to the choice of the analysis/business question being studied which in turn sometimes

leads to relevant findings. Combined they create a realization of analytics-ready data that is

high-quality and factual to execute precise and informative analytics, which are useful to

organizations in development of business values.

Week 2 Assignment

Exercise

Question 12

The dataset found on Data. gov under the title "Electric Vehicle Population Data,"

offers complete information about the Battery Electric Vehicles (BEVs) and Plug-In Hybrid

Electric Vehicles (PHEVs) number as registered with the Washington State Department of

Licensing (DOL). Created on April 19, 2024, the data which is hereinabove contains items

like make, model, and type of electric vehicles (Electric Vehicle Population Data Sheets,

2024).

This data set underlie a chart which portrays the leading ten cities in Washington State

that have numerous electric vehicles. Through the indexing of electric vehicles per city, an

individual profile is provided conveying how the distribution of the manufacturers' electric

cars in these cities is portrayed. Cities ranked in the top ten included Seattle, Olympia, Lacey,

Tenino, and Yakima, and several others did but may not have been included in the list. The

selected cities were identified as those that most of the electric vehicles registrations occur;

an apparent concentration of the electric vehicle adoption among the population in these

cities.
6

Week 2 Assignment

A categorical bar chart was constructed, as an alternative way to represent visually

this data set. The graph demarked each of the cities as a bar fashioned with a brand name of

the electric vehicles being sold. At first, the dash showed various makers from electric cars,

so there was a mass of data on the chart. For better clarity, it has been decided to present the

chart only with the most ten electrical cars brands. This was the more commendable zcn. It

was less zoomed, thus easier to point out the distribution trends of the most sought-after

electric vehicle brands in the top ten cities.

Besides this dataset, there are also literature including instruments and service region

design that contribute to the explanation of the observed trends accordingly. They depict that

era of urban electric vehicle sharing systems is absolutely about service region design

efficiency and users’ acceptance (2017; 651). Study “Improving Accessibility and

Convenience of the Electric Vehicle Services” by Zugurova and Krasnopeytova was

published in the Manufacturing & Service Operations Management journal and the focus is

strategic planning for mounting the electric vehicle services in urban areas. This observation

comes along with the fact that advanced cities are the most significant locations for the pure

electric vehicles in the world and thereof proves that urban infrastructure and planning are the

vital elements in the EV adoption.

Chapter 4

Discussion Question

Question 1

The data mining is the operation of data discovering meaningful information which is

hidden in large datasets by using scientific statistics, machine learning, and computations

techniques. It is all about taking the huge batch of information stored in databases, data

warehouses or other data containers and through it, extracts the most important knowledge. It
7

Week 2 Assignment

is a targeted process – processing unprocessed data into valuable insights that inform

decision-making and strategic planning (Marconi et al. , 2019). The data mining processes

encompass revealing hidden patterns, which, in turn, help the organizations to predict trends

for the future, improve processes within the organization, and make the strategy development

more effective.

Data mining works at the junction of numerous fields, which makes it impossible to

name this notion. These fields are statistics, computer science, and artificial intelligence.

Through different fields great attention is paid to some special aspects of data mining, so that

as a result variety of terminologies and perspectives are demonstrated. (Marconi, et al.

(2019))Furthermore, as data mining methods and applications get updated through the

emerging techniques and definitions, the methodological variations happen to be captured in

the new methods. Diversity of the teams shows the richness of different applications, the

great promise and the constant development of data mining as a field. In addition to that, the

various terms that are used are an indication of the interdisciplinary nature of data mining as

well as the flexibility of this field of study to tackling unique problems of different context

and industries.

Question 2

The recent introduction of data mining as an integral part of business activities can be

explained by a number of crucial factors. Initially, big data caused the rapid expansion of data

created by enterprises, social media and general Internet of Things (IoT) devices, which

overwhelm conventional analytical tools. Data mining techniques allow companies to extract

meaningful information from the ocean of these big datasets. This knowledge helps them to

make solid decisions and obtain leading prospects over rivals. Things that can be added to
8

Week 2 Assignment

this is that, thanks to computational power and machine learning algorithms, data mining has

become more advanced and efficient, hence making it more widely used.

Furthermore, the growing requirement for data driven decision making in multiple

fields resulted into the popularity of data mining. Organizations target reducing redundancy

in their processes, ensuring better consumer satisfaction and finding unseen prospects with

data analytics (Marconi et al. , 2019). Predictive and prescriptive capabilities forecast trends,

detect abnormal behavior, and identify ideal solutions to improve strategic decision-making.

In addition, both rules and controls contribute to the requirement for high standards of data

analysis thus relates data mining as a necessity for operation of a modern business.

Question 3

Prior to procurement of data mining software, an organization must evaluate what or

for what its particular needs and objectives are. Identifying the purpose, like realizing clients’

spheres or optimizing operational operations, will contribute to the picking of needed

instruments (Marconi et al. , 2019). Further of it, the organization also must consider the

scalability software enabling it to take over large amount of data and the ability to exist along

other systems.

Besides simplicity and difficulty of learning that comprise a software, another vital

one is also a usability. For instance, organizations need to evaluate from time to time if their

team possess the sufficient talents or some training is needed (Marconi et al. , 2019).

Moreover, support and documentation offered in the software and the vendor's good customer

help can affect the implementation of the data mining projects positively and also the long

term success of the projects. One should also compare costs associated with its use (i. e.

licensing feel and other hidden costs) with the efficiency of investment.
9

Week 2 Assignment

Question 4

Data mining, unlike the other analytical instruments, can automatically unveil the

hidden patterns and the relationships which exist within the large datasets. Different from the

conventional statistical analysis, which primarily tests hypotheses that are predefined, the

data mining puts algorithms to work against data, coming up with hypotheses (Marconi et

al. , 2019). Such investigative capacity makes data mining a remarkable instrument for

discovering of patterns and regroups that are invisible in conventional research. Data mining

equally incorporates disciplines like machine learning and artificial intelligence and includes

a full-scale analytical approach.

Moreover, data mining concentrates on predictive analytics that is used to anticipate

future tendencies and patterns whilst business intelligence is all about descriptive analytics

which is mainly used to summarize historical data. This ability to predict the future leads

organization to take proactive and data - driven decisions which they cannot with traditional

analytical techniques that provide mostly retrospective insights. Due to the multi-dimensional

and deep nature of data mining, it is extremely efficient in pursuit of a very complex pattern

and application.

Question 5

Data mining methods that are most prevalent may be divided into four categories -

classification, clustering, association rule mining, and regression. To classify data, predefined

categories area chosen based on similar patterns learned from a given training dataset

(Marconi et al. , 2019). Clustering acts as a mechanism that clusters groups of data points that

share common traits while being ignorant of the imposed labels, thus illuminating the true

structures of the data. The association rule mining technique gives frequencies in variables in

large data sets showing frequent item sets and correlations. Regression as we know it is the
10

Week 2 Assignment

techniques to predict continuous outcomes from variables which are mapped to one another

as dependents and independents.

Principally, the two approaches differ in the basis and tactics that they involve.

Classification is concerned with the prediction of the categories of variables whereas

clustering is towards segregating data without using any predefined labels (Marconi et al. ,

2019). The purpose of the variable is the association rule mining which endeavors to uncover

co-appearances and co-occurrences sequentially and for the predictive model. Meanwhile,

such differences allow both techniques to take part in different types of data analysis tasks.

They are consequently important in the toolkit of the data miner.

Exercise

Question 1

The latest in data mining and predictive modeling has taken great strides, most

particularly, with GANs' incorporation into advanced schemes like Generative Adversarial

Networks (GANs). Official release of institute papers published by Teradata University

Network reveal a growing trend towards factoring in causal patterns that contribute to

increased accuracy of machine learning. A use of conventional techniques heavily relies on

correlation, which does not provide full data picture with the distribution shifts between

training and testing data. Whereas the budget restraint is tackled by the integration of causal

information into the predictive models in order to raise the robustness and fitness for purpose

of the machine learning models (Zeng et al. , 2024).

One major advance they proposed is the causal information learning GAN-based

framework which Zeng et al, (2024) introduced. This novel method, based on the

relationships of the causes and tasks in everyday objects such as body movement and health,

is used for sensing non-linear datasets. By paying attention to certain at-risk variables and
11

Week 2 Assignment

removing confounding bias systematically, such a system improves the accuracy of the

modern deep learning technology that already works. Extensive tests performed in domains

relevant to large-scale and are the best argument for these in comparison with existing models

ranking high (Zeng et al. , 2024).

Such developments are not confined to just one sector; but rather, they cut across

many fields like health and mobility that have high demands for precise predictions. Along

with the capability to manage the distribution shifts and the causal data through advanced

analytics which is indicative of a huge leap in the field of predictive analytics. Beyond their

contribution of demonstrating the superiority of GANs approach for the prediction of

disaster-related air pollutants, Zeng et al. (2024) also highlight the need for more research in

this field to identify and assess any other possible improvements in this technique. Therefore,

the combination of causal explanations with machine learning expert models is one of the

major developments envisaged at this time for increasing the reliability and use of predictive

analysis.
12

Week 2 Assignment

References
Electric Vehicle Population Data. (2024, April 19). Retrieved from

https://catalog.data.gov/dataset/electric-vehicle-population-data

He, L., Mak, H.-Y., Rong, Y., & Shen, Z.-J. M. (2017). Service Region Design for Urban

Electric Vehicle Sharing Systems. Manufacturing & Service Operations Management,

19(2), 309–327. https://doi.org/10.1287/msom.2016.0611

Klee, S., Janson, A., & Leimeister, J. M. (2021). How Data Analytics Competencies Can

Foster Business Value- A Systematic Review and Way Forward. Information Systems

Management, 38(3), 200–217. https://doi.org/10.1080/10580530.2021.1894515

Wo, J. C. (2019). Revisiting the Crime Control Benefits of Voluntary Organizations:

Organizational Presence, Organizational Capacity, and Crime Rates in Los Angeles

Neighborhoods. Crime and Delinquency, 65(7), 916–940.

https://doi.org/10.1177/0011128718787517

Zeng, J., Zhang, G., Yuan, J., Li, Y., & Jin, D. (2024). Empowering predictive modeling by

GAN-based causal information learning. ACM Transactions on Intelligent Systems

and Technology, 15(3), Article 54, 19 pages. https://doi.org/10.1145/3652610