Data Mining Techniques and Applications

This assignment delves into the realm of data mining, examining key concepts and techniques such as association rule mining and neural networks. It highlights the diverse applications of data mining across industries including healthcare, finance, and manufacturing. The document explores how these techniques are used to uncover patterns, trends, and insights from vast datasets, enabling informed decision-making and business growth.

TASK A
Question 1:
Define and explain the following terminologies:
 Just-in-Time Delivery
The system Just-in-Time has been stated and establishes across the world by many
practitioners and researchers for the task of manufacturing. So, Just-in-Time is a listing
strategy for different corporations use to extend potency and minimize was with the help of
receiving goods as they are required within the production method by that minimizing
inventory prices. So, forecast demand exactly is needed by this technique. So, that the
product reaches on time at delivery point.
 E-Commerce
E-commerce is that the shopping for and mercantilism of products and services, or sending
knowledge or funds done with an electronic system, basically the web. So, all these business
transactions take place either as consumer-to-consumer, business-to-business, business-to-
consumer or consumer-to-business. Typically, the terminology e-commerce and e-business
are used conversely. The terminology e-tail is additionally generally employed regarding
transactional process for line looking.
 SaaS
Delivering the applications done with the internet as a service is done by SaaS (Software as a
service), so we can say that SaaS is the means of delivering applications, comparatively
putting inside and then maintaining computer code, you just approach it with the help of the
web by releasing yourself from advanced computer code and hardware management.
SaaS application are generally known as Web-based Computer code, on-demand computer
code, or hosted computer code. Behindhand of this name, all the SaaS applications run on a
SaaS provider’s servers. The supplier managers approach to the applying, in sync with
security, convince and performance.
 Strategic planning
Strategic planning or designing is associate structure management activity that is worn to set
preferences which focuses energy and assets, enhance operations, make sure that staff and
alternative stakeholders area unit operating towards familiar goals, build mediation around
supposed outcomes/results and admittance and change the corporation’s guidance in reply to

a dynamically surroundings. It is a disciplined effort that generated basic choices and
activity that form and supervise what a company is, an agency it serves, what it will and why
will it with full attention on the long run. Useful strategic planning/designing express not
solely wherever a company goes and the activity required to from development, however
additionally however it will grasp if it is made.
 Supply Chain systems
Supply Chain System helps commodity firms to make superior provide chains. We tend apply
operations, capital and management all told areas, exploitation leading techniques and world
power. SCS experts all have decades of direct, active expertise – we all know a way to apply
performance and the way to induce things done. We tend to attempt to superior consumer
service, honesty and price.
 DSS
A Decision Support System (DSS) could be a bug application which analyses business
information and represent it in order that users will create business selections additional
simply. It is an “information application” (to differentiate it from “operational application”
which gather the information within the course of traditional business operation).
 Cloud infrastructure
In IT domain where we attribute to the software and hardware components – like servers,
network, storage and virtualization package – which square measure required to device the
computing needs of a cloud computing model called as cloud infrastructure.
 Web 2.0
Web 2.0 is that the present position of on-line technology because it correlates to the
first days of the net, characterised by larger user interactivity and
combination, additional common network property and increased communication channels.
One among the foremost vital variations in between internet a pair 2.0 and therefore
the ancient World-Wide internet (WWW, retroactively mentioned as internet 1.0) is
bigger collaboration in between web users, content suppliers and enterprises.
Basically, information was announcing on internet sites, and users merely viewed or

downloaded the content. more and more, users have additional input into the character and
capacity of website and in some cases, strive period management over it.
 Extranet
Related to a company computer network, associate Extranet may be a native network
available to outdoor users or businesses, however closed to inaccessible to the overall public.
Though a user might acquire access to associate extranet, the extranet might solely show
information that he or she is authorized to ascertain or solely enable access to limited part of
the extranet.
 Big data analysis
Big Data Analysis is that the method of criticizing massive and divers data sets – i.e., massive
data/knowledge – to make known hidden patterns, unexplored correlation, market trends,
client desire and alternative helpful info which will facilitate organization create more-
learned business choices.

Question 2:
What are the key elements of Data Mining? Discuss with Examples.
Answer:
Data Types: Data mining can work on any type of data like dimensional databases and
relational and Structured data in spread sheets etc.
Analysis Style: It is focused on forecasting and exploration of significant business factors on
little scale victimization process intelligence. Which needs management professionals
operating in shut association with information individual.
Expected Results: When a user needs critical decision making than data mining gives result
in report with recommendation.
Focus: Data mining helps in generating new analysis indications for BI and intensify data
patterns.
Data volume: For working with very high data processing costs data can work on small
datasets and sample (small part) data also.
Examples: Data mining can be used anywhere we have massive amount of data like –
Business, Science and Engineering, Spatial data mining, temporal data mining, medical data
mining, human rights, music data mining, pattern mining etc. These are very few example of
data mining.

Question 3:
Why understanding the significance of “Reserve Stock Level’’ function of ERP is so
critical for the systems performance? Explain your answer.
Answer: The “Reserve Stock Level” is one of challenging topic in ERP. Reserving stocks are
primarily worn to make sure the stock availableness once delivering a sales order. Returning
with this characteristic it is additionally attainable to lock stock not associated with any sales
order to avoid consumption. A reservation indicates sure stock within the warehouse that is
reserved and cannot be consumed by anyone at all apart from the owner of the reservation.
Presently there are a pair of attainable home owners, a sale order line or the system. A system
reservation may be an appropriate may be a special reservation kind that can’t be utilize by
anyone. System reservations are worn for dominance reservation once some stock must be
blocked within the warehouse. 2 sorts of reservations come as practically developed:
Pre-Reservation: These are reservations that do not seem to be physically within the
warehouse however ordered to a provider and whenever there is a relationship between the
road of the acquisition order and a sales order line. Once the acquisition order line is
acknowledged this pre-reservation is mechanically born-again to a reservation.
Reservation: It refers to stock keep within the warehouse that is already reserved by a sales
order line.
A reservation is often outlined by the merchandise that is appetite to be reserved however
different proportion will be outlined like the warehouse, storage bin and attribute. Another
attention-grabbing issue is that the chance of allocating or not the reservation:
 Allocating the stock means definite stock is reserved for a sales order, instead of it
being a general reservation. That exact stock cannot be reserved for the other sales
order.
 A not allowed reserved stock will be modified at any time by different existing stock
however invariably guaranteeing that the sales order keeps the reservation.
These are the main reason we say that reserve stock level is the most difficult function of
ERP. As it needs to pass out some more function before doing final action.

Task B
SWOT Analysis for RealEstate.com.au
Answer:
In our organization that is RealEstate.com.au we need to update or change is customer
support system at any point of time with any type of service/facility etc. Should also start
giving shifting and settle down help to customer so that does not need to find another vendor
for setting up the house. But we must figure out SWOT according to our requirement which
is listed down below:
Strengths
 Multiple services
 Easy to search
 Continuous
customer support
 Security for all
Weaknesses
 Avoid only profit
making
 Keep key features to
compare with
competitor
 Making customer
profitable
Opportunities
 Automated
Services
 Government
authorization
 24*7 support for
making more
customers
 Improving
Lifestyle
Threats
 Recent Technology
 Govt. regulations
 Competitor
reducing prices
 Product quality
 Customer not
satisfied
SWOT

These can be identified once we define our requirement for updating our business. As
strength of our business is the multiple services with customer’s custom choice. And the
biggest threat is that our competitor continuously reducing the prices than we need to analyse
what we can do to gain more customers. SO, if we add one or two more services within the
same price than we can’t lose our customer. Because making loyal customer is one of the
opportunity for our organization. As students usually looks for cheap and best services, so
there must be special discount or multiple options for all type of students depends on their
budget.
Short Term: Customizing the service of stay according to the customer. If want to stay for
shorter duration than also provide good services to them.
Mid Term: Management and analysis of our database.
Long Term: First, we will analyse our complete data like what type of customers do we have
more, maximum budget service taken by how many customers, which service customer
taking more etc. Because analysis will help us to define long term gaols. Like if we have
more number of students in our customer list than we will plan to increase some facilities and
will plan out to gain large hour/apartment etc customers.

Task C
Paper on “Data Mining”
ABSTRACT
Data mining could be a method that finds helpful patterns from great deal of knowledge. This
paper talks about few of the info mining algorithms, techniques and a
few companies that pickup custom-made data processing techniques to enhance their
businesses and located wonderful results. In this paper, the construct of knowledge mining
was summarized and its implication approaching its methodologies was explained. The data
mining supported Neural Network and Genetic algorithmic program is researched very
well and the main technology and ways that to realize the info
mining on Neural Network and Genetic algorithmic program are surveyed.
I. Introduction
Data Mining introduces to separate out or mining the information form huge amount of data.
The basic term data processing is suitably known as ‘knowledge mining from data’ or
‘knowledge mining’. The storage and assortment technology of data has created it potential
for companies to acquire immense amount of knowledge at lower cost. Manipulating this
hold on information, to extract helpful and actionable data, is that the complete objective of
the universal activity named as data processing. The subsequent definition is given: An
approach for examining and analysing with automatic or semiautomatic suggests that, huge
amount of knowledge to get purposeful patterns and rules known as Data Mining. The recent
evolution of information technology has cause great bulk of databases and also big data in
heterogenous domain. The overall analysis in data and databases technology has inclined
towards associate in nursing path to store and change this inestimable information for
additional deciding. Information processing may be an action of extraction of appropriate
info and patterns from large amount of information.
It is additionally referred to as an information discovery method, knowledge
mining from huge set of information, extracting information or information /analysing

pattern. Data mining may be a cogent and logical method which is accustomed finding with
great extent of information to search out appropriate information. The objective of such
method is about to search out patterns which were antecedent unknown. But once all such
patterns square measure initiates they will additional be accustomed make sure selections for
advancement of their businesses. There are basically 3 steps which are involved in data
mining activity:
 Exploration
 Pattern Recognition
 Deployment/Readying
Exploration: Within the start about data exploration/analysis first data is cleansed and then
reworked into different type and meaningful variables so quality of information supported the
matter square measured determined.
Pattern Recognition: After data exploration step takes place, retrained and outlined purpose
of the specialized variables the second track is to create pattern recognition. Determine and
opt for that pattern that build the most effective prediction.
Deployment/Readying: Patterns square measure set up for desired outcome.
II. Techniques of Data Mining
There are heterogenous techniques are being used according to data user have and needed
result. So, user can use any of the technique like Clustering, Classification, Regression,
Association Rules, Neural Networks, Decision Trees, Genetic Algorithms, Artificial
Intelligence, Nearest Neighbour etc, can be use by user for knowledge discovery from
databases. Classification
The uttermost ordinarily practiced data processing technique is known as classification.
Which apply a group of re-classified prototype towards advance a model which can classify
the populace of massive data.
For an example detecting frauds and risk function for credit are notably compatible
towards presenting the variety of analysis. This way ofttimes apply neural networks or

decision tree positioned classification techniques and algorithms. The information
classification method associates learning and then classification.
In the learning phase, coaching knowledge are being analysed with any of the
classification algorithmic program. The exactness of the classification rules is evaluated by
classification check knowledge. In case if the exactness is adequate the foundations may be
enforced towards the new knowledge tuples. Like in the example of detecting fraud
application, it will consider full records of each dishonest and authentic actions found on a
record-to-record base. Now, the training part of classifier with algorithm program which uses
these re-classified examples to work out the prescribed specification needed for correct
inequity. The algorithmic program then put into code these specifications to a model which is
known as classifier.
List of different classification models:
 Classification by decision tree induction
 Support Vector Machines (SVM)
 Bayesian Classification
 Classification Based on Associations
 Neural Networks
 Clustering
In the list of unsupervised learning, clustering is the most important problem which is used to
aforesaid as identifying the comparable categories of various objects. By victimisation bunch
approach we will in addition to establish close and distributed fields in object area and may
come upon comprehensive dissemination trends and parallel with knowledge characteristics.
The classification technique advance may be worn for useful factor that of differentiating
teams or categories of item however it becomes pricey therefore bunch may be used as pre-
processing way to aspect set choice, classification. As an example for making cluster from
consumer supported purchasing pattern, to classes genes along identical practicality.
List of clustering techniques:
 Partitioning Methods
 Hierarchical Agglomerative (divisive) methods

 Density based methods
 Grid-based methods
 Model-based method
 Prediction
Different technique of regression may be tailored considering declaration. Analysis using
Multivariate may be worn to miniature the communication within one or a lot of freelance
variables, vulnerable variables. In the step of data processing, autonomous variables are
aspect as of now identified and most of the time what we would like to forecast is known as
response variables. Unsuccessfully, several physical world issues don’t seem to be merely
prediction.
As an example, stock prices, sales volumes and merchandise breakdown rated are
complete terribly troublesome to forecast as a result of they will rely on advanced
communication about many forecast variables. So, a lot of advanced techniques could also be
important to estimate future values. A similar model variety will typically be worn for each
classification and regression. As an example, the CART (Classification and Regression
Trees) call tree algorithmic program may be worn to create each tree of classification and
regression trees. Neural networks can also produce one and the other regression and
classification models.
There are some regression which are listed below:
 Nonlinear Regression
 Linear Regression
 Multivariate Nonlinear Regression
 Multivariate Linear Regression
 Association Rule
Correlation along with association are typically into seeking out concurrent information
group decision within giant information sets. So, this kind of discovery support business to
create sure choices, like catalogue style, cross promoting and client shopping behaviour

reasoning/analysis. Association Rule logic got anticipated capable into producing guidelines
confidently values but one. But the amount of doable association rules as accustomed data
sheet is mostly appropriate giant with a huge capacity of the principles are sometimes of very
little worth.
Various categories of association rule
 Multidimensional association rule
 Multilevel association rule
 Quantitative association rule
 Neural Networks
Group of interconnected information input, output entities and every association includes
weight gift with it and this is known as neural networks. At the time the educational section,
network determine/learns by accustom weights thus on can predict the right category stamp
of the input tuples. So, neural networks have the exceptional capability towards evolving
which means in distinction too difficult as a choice general information along with might be
accustomed excerpt patterns and sight trends that are too advanced to be detected by either
humans or alternative pc techniques. These are like minded as stable valued inputs and
outputs. If instance, written character acceptance considering coaching pc to utter English
text several world business issues and have already been with success enforced in many
industries. So, neural networks are the best at distinctive trends in information, likeminded
for forecasting or estimating wants.
Neural network can be of any type:
 Back Propagation
 Forward Propagation
III. Data Mining Applications
 Future Healthcare
 Market Basket Analysis
 Manufacturing Engineering
 Education
 Fraud Detection

 Intrusion Detection
 Customer Segmentation
 Financial Banking
 Research Analysis
 Bio Informatics
 Criminal Investigation
There are substantial number of organization which are using Data Mining for their benefits
to make data useful. Some of the organizations are listed below:
 FBTO Dutch Insurance Company
 ECtel Ltd., Israel
 Provident Financials’ Home Credit Division, United Kingdom
 Standard Life Mutual Financial Services Companies
 Shenandoah Life insurance company United States.
 Soft map Company Ltd., Tokyo
IV. Conclusion
Data mining has very high priority related to discovering the trends, statement, analysis of
data etc., in several business spheres. Data processing techniques are like classification,
neural networks, clustering etc. benefit to find out trends/pattern toward make your mind
up upon the long run directions in businesses to get bigger. Data processing hold ample
application domain virtually in each business wherever effective information in produced
so that is why data processing taken each in every of the foremost vital bound within info,
knowledge systems and each in every of foremost encouraging knowledge domain
evolutions in information technology.

REFERENCES
Jiawei Han and Micheline Kamber. (2004). Data Mining Concepts and Techniques.USA
Dr. Gary Parker. (2004). Data Mining: Modules in emerging fields. CD-ROM.
Customer Successes in your industry from. Retrieved from
http://www.spss.com/success/?source=homepage&hpzone=nav_bar.