American Journal of Engineering Research (AJER)

2016
American Journal of Engineering Research (AJER)
e-ISSN: 2320-0847 p-ISSN : 2320-0936
Volume-5, Issue-9, pp-09-14
www.ajer.org
Research Paper
Open Access

Data Mining and Decision MakingApplied Model

KhashmAlgerba Dam in Estrin Sudan
Dr. Alshafie Gafaar Mhmoud Mohmmed
Computer Science Department,College of Science and human studies Al-Ghat,
Majmaah University, Saudi Arabia
Prof. Saife Eldin Fatoh Osman, Dean of Emirates Collage of Science &
Technology, Sudan
[email protected]

Email:[email protected]

ABSTRACT: The use of Technical Data Mining enables discovering and concentrating on the most important
information of database in all fields. It also leads to future predictions, explore behaviors, and identify
directions that ensure taking decisions effectively. This paper will shed some light on data mining, and the
extent to which it contributes to taking decisions effectively.

I. INTRODUCTION
The current era is distinguished with sophisticated technological revolution such as the Internet and
Digital Economics. The flood of data and its rapid development suggests that it is difficult and even impossible
for an analyzer to extract valuable information by using classical entrances to analyze primary data, especially
with a large amount of data stored in databases and storage devices. There is thus an urgent need for developing
storage tools, data analysis software, in addition to increasing extracting information and knowledge.

Figure1.Source:http://blog.naseej.com
Consequently, it appears that data Mining is able to extract information from large databases. This new
modern technology is considered a breakthrough in information revolution. It helps companies and
organizations discover and concentrate on the most important database. Mining Technology concentrates on
making accurate future predictions and discover behaviors and aptitudes, which lead to taking appropriate
decisions in a suitable time. This technology is however more complicated when used for discovering
information in databases. It is related to a large extend to the operation of another development, which is very
important in data storage. Most of the companies nowadays tend to discover information systematically and
methodologically, which is considered as a backup that works best by activating effective systems to achieve
distinctive competitions.

www.ajer.org

Page 9

American Journal of Engineering Research (AJER)

2016

First:The concept of Data Base Mining:

The concept data Mining emerged in the mid of the nineteenth century in the United States of America.
It includes both static and technology (Database, artificial intelligence & Machine learning). There are several
yet interrelated definitions of data mining. One of these definitions is the automated discovery of common
pattern hidden database" (Cite). It also refers to the procedures of intellect precise interactive analyses and
series for runners of activities when using these procedures in decision taking and making appropriate activities
on the benefit of the task that they are responsible for and the institution where they work (Cite). Another
similar definition is "it is the analysis of a large amount of data so as to find bases examples and models which
used as guides for decision makers, and predict future behavior" (Cite). Moreover, it refers to the analysis of a
large size of data for the search of possible relation and summarize data in a new forms in order to be
comprehensible and beneficial for its users" (Cite).
Based on the aforementioned definitions, one may realize that that Data Mining aims at extracting
hidden information in databases. It helps institutions discover and focus on significant information found in
large-scale databases. Such technology has become a research interest for many researchers to take advantage
of. Furthermore, the rapid increase of data warehouses has made it necessary to explore effective alternative
techniques for extracting information and knowledge from this crowded data, which are used for problem
solving and decisions taking using modern computer applications. This form of data extraction is a new
sophisticated technology for making computers act as human beings, perform as human workers. According to
(Cite), it refers to "the idea of the discovering and mining data with sophisticated methods in problem solving
and decision taking processes (Cite). Moreover, this concept is considered as a fundamental step of
knowledge discovering by data baser (Cite).
Second:Factors of Data Mining Revolution
Factors of data Mining revolution can be divided into two main factors:
1. Predictive Model.
2. Descriptive Model.

Figure 2

II. DATA MINING

Types and tasks of Data Mining Technology
There are two models of Data Mining:
1. Predictive Models
2. Descriptive Models
The predictive Models help predict the value of some characteristics such as the predictions of water
levels time series. Time series can be defined as a phenomenon value in a series of historical sequence days,
months and years. The purpose of studying time series is to analyze a set of these series and the accompanying
changes made by the phenomenon, which takes place through a period of time. It also aims to analyze the
causes, results, and predictions depending on an idea that took place in the past and continued to the future.
Another reason for studying time series is to make use of its analyses so as to predict the future.
To perform this analysis effectively, a growth scale, which interprets a mathematically, should be used
so that the growth grade of time series, which is the purpose of the study, can be limited. Moreover, the general
deviation of these series can be determined in addition to limiting its value that leads to the future.
Youssef (2002) points out that "a group of models are based on a hypothesis which proposes that the
behavior of a future phenomenon depends entirely on its past behavior. Similarly, the prediction value also
depends substantially on the phenomenon value in the past. For example, one of the widely recognized models
is Jenkiz Box Model, which depends on a certain time series. However, interpretative variables, such as time
series models and the falling models, are used for making perditions. These two models take the following form:

Z t f ( xt , )

www.ajer.org

Page 10

American Journal of Engineering Research (AJER)

2016

Z t represents a variable used for predicting in the future.

xt represents the independent model or interpretative model, which functions as a cursor of time (e.g.
Trigonometric functions polynomial functions).
represents parameters used for estimations. These parameters should remain in downhill models and also in
series time model analysis even though this hypothesis has been deleted.
represents the error of estimation.
The descriptive Method has been divided into two categories: 1- Held models which permit the assembly of individuals, actions or products in clusters.
2- Creative Models which permit determination of relationships.
There are many tools for Mining in Data such as: 1- Summarization: It refers to the method that transforms huge data into summarized matrices that provide a general
description of the variables and their relations such as averages, totals, and statistical descriptions, which help
generate the mean, median, mode, standard deviation. Although these standards provide large summaries, they
often neglect details related to the consumer behavior.
2-Classification: Classification represents the interpretation or prediction according to individual specialization through
other characteristics that can be undertaken through conducting classical statistics such as downhill and
classification analyses or depending on modern methods such as link forces, depending on solutions, or neural
networks.Examples of classification methods can be seen in the application of knowledge discovery, which
includes directions of financial markets and electronic determination of important things (e.g. Database).
3-Prediction: It deals with estimation or classification processes regardless of how data have been classified based on
future prediction methods. Prediction is used for determining future values of the dependent variables
quantitatively. Some of the traditional tools used in prediction processes are downhill and distinctive analyses.
However, there are more effective and accurate modern methods including but not limited to linking forces,
decision tree, neural networking, and inherited algorithms.
4-Clustering
It classifies the population data into homogenous groups, representing individuals, families, or friends
depending on the information included in the total variable. The purpose of this clustering is to develop
marketing programs designed in an average of customers themselves. The flakiness assembly method is used in
most cases by the aid of statistical flakiness method or by those of the decision tree, the neural networking, or
internet algorithms.
5-Rule Analysis:Rule represented by the search for relations or links exist between different characteristics. Nonetheless,
rule analysis refers to a group of methods used for linking purchase patterns with sector or through time (e.g. the
use of the analysis used in market baskets to get benefit of the potential information of goods purchased by
actual customers to predict the goods actually purchased if there is especial offer for them or if they have been
informed by the goods).
6-Dection change and deviation: It depends on the most discovered variables of data through previous majors or standard values.
Fourth:Levels of Data Mining process:
Levels and steps of data Mining process can be summarized as follow:A-Business understanding is the first issue of knowledge discovery that deals with understanding the problems
that face business. For example, how can the greatest beneficial of data mining be achieved which require a
clear limited formula of business goals.
B-Data Understanding

www.ajer.org

Page 11

American Journal of Engineering Research (AJER)

2016

The understanding of the nature of data is the most important factor in data mining processes and
knowledge discovery for gaining a deep understanding of the given data. It refers to the help of designers to use
tools Algorithms used in limited precise issues. This provides great opportunities for success in addition to
raising the efficiency and qualification of knowledge discovery systems.
The most essential steps of understanding data process can be summarized as follow: 1- Data collection: It deals with limiting sources of data used in the study included over general data such as taxes.
2- Data description: It focuses on the description and contents of files or charts.
3- Explanatory analysis of data;It uses the photography direct analysis method, which leads to the Primary analysis of data. This point
is crucial, as it is based on the development of hypothesis related to the problem of the study.
Data preparation:It includes the following:
A. Selection: refers to the expected variables collection and sampling.
B. Variable construction and transformation: is an essential process for contextualizing new variables to
construct effective models.
C. Data Integration: The collection of data in data Mining studies is considered here and database with multipurpose in only one
database is also taken into account.
D. Data formation and designing: This step is concerned with the formation and order of data fields, as it is required in the model of data mining.
1-Model Building and validation: building a precise solution model takes place through the process of trial and
error. In most cases, this process requires an aid from specialized individual in data mining so as to select and
check different alternatives for achieving the best models that can be solve the problems under investigation.
5- Evaluation and Interpretation of findings.
When the model is contextualized and its validity and reliability are en/*sured, then the process of
confirming data reliability, which fed by the model, is initiated. Once the outcome of this data is clear, then the
expected findings can be compared with the actual data under study in order to reach the precise model.
Model Deployment:This stage includes developing an organizational model that helps take decisions. The valid model
should then achieve users satisfaction. However, the model can be selected through directed study or a simple
sample model from a population sample.
Applied Frame

Figure3- A chart explains the downhill model series of a wasted water I a channel.
explains the data time series for the wasted level in the channel.

www.ajer.org

Page 12

American Journal of Engineering Research (AJER)

2016

Figure 4: Based on the chart series above, it seems that there is no general direction. This means that
the series was stable in the middle. Moreover, we noticed that there is no change in the series.
The stability test of a wasted series in a channel.
Dis_ch
AR1
constant

B
.7304185
4.3307376

SEB
.03788795
.27596051

T-Ratio
19.278387
15.693324

Approx.prob
.0000000
.0000000

Residual Variance: 1.8231508

Chart 3 Explains the data time series for a level of a wasted channel
Figure 3
Based on the chart above, we can notice all the corelated efficiency located inside the granted period,
which refers to the constant time series (Random data).
A model of downhill series of a wasted water in a channel

X t 4.3307 0.7304 X t -1
Explains the Downhill Model from Selected class of a Series Wasted Water in a Channel.

Figure 5
Explains the prediction of a channel wasted level.
Expected figures of a main wasted water in a channel
We notice that the expected figures had increased, which influences the irrigation of lands.
Consequently, it reflects on the production of the scheme.

III. CONCLUSION
The mixture of information technology and statistical method leads to providing appropriate suitable
prediction of future behaviors. Then, it seeks for suitable solutions for the problems before they occur. If it
exists, the prediction of general and refresh in many domains using the technical of data mining, which forms a
level of a whole knowledge discovery in database. In sum, it becomes one of the most important issues that
make countries engage their institutions.

BIBLIOGRAPHY
[1]
[2]

Bashir A bbass .Digital Administration in Field and A applications . Emirates Centre of Studies and Research straggles .Abu Dubi
.2005.Pages.84 .91-89.92.93.93-94.
Bazsalica M.Naim.p.Data Mining pour Le Web edEyrolles .Paris .2001.pages.01.68-69.66.68.

www.ajer.org

Page 13

American Journal of Engineering Research (AJER)

[3]
[4]

2016

AbdAl-star Alali ,Abbass IBRAHIM Gedanhaly . GassanALamri. Introduction to KMOWLEDGE Administration .DarrAlmsyra for
publishing , printing and distribution. First Edition . Oman pages 177.100-102 .99-98.
Berry J.A.M Linoff.g.s Data Mining Techniques for Marketing Sales and Customer relationship Management 2nd Edition .Wiley
Publishing INC.indianapolis.2014 .P .10

www.ajer.org

Page 14