prediction using data mining.pdf

All Rights Reserved © 2019 IJERCSE 1
ISSN (Online) 2394-2320
International Journal of Engineering Research in Computer Science and Engineering
(IJERCSE)
Vol 6, Issue 1, January 2019
Prediction Using Data Mining Techniques and
Tools
A.Sahaya Arthy
Research Scholar, Department of Computer Science
Manonmaniam Sundaranar University, Tirunelveli, Tamilnadu, India
Abstract: - Data mining is a process of extracting knowledge by analyzing various past ancient data bases and predict the end result
or take appropriate decision based on prediction. This data mining process has many challenging issues while performing research,
analyzing raw data of the past records and predicting may also lead to negative decision result few times. Direct application of
methods and techniques developed under related studies in machine learning, statistics and database systems cannot solve these
problem. It is required to perform dedicated and appropriate analytical studies to invent new data mining methods or to develop
integrated unique techniques for efficient and effective data mining, whereas data mining itself has formed an independent unique
and innovative field of study. Data mining is a widely used platform to perform various decisions making in many Industry like
banking, finance, agriculture, communication, telecom, Military Service, Police department, other government departments,
engineering, Medication &hospitals, law & order etc. This paper deals with detailed study of Data Mining, its techniques, tasks and
related tools.
Keywords: - Data mining Techniques; Data mining algorithms; Data mining tools.
I. INTRODUCTION
Data mining is an extraction of hidden predictive
information and acquires knowledge by analyzing a large
database. Technically, data mining is the process of
discovering correlations or patterns among large number
of fields from relational databases. It is a strong tool used
widely because it can provide with relevant information
that anyone can use for their own advantage. When we
have the right data, then we have to apply it in the right
manner and then we will be able to get beneficial
appropriate result. Now a day’s getting information is
relatively easy as we have many sources to grab the
relevant data. But to achieve certain desired goals it is
required to get the most relevant information. This is
where data mining becomes a powerful tool that we will
become very familiar with. Data mining gives us the
power to predict certain behaviors within a system. Data
mining involves the high standard detection, association,
regression, rule learning, classification, and
summarization and clustering. The ultimate goal of this
technique is to find various patterns that were previously
unknown. Hence the mined results should be valid,
novel, useful, and understandable. Once we find these
patterns, we can use it to solve many problems.
Figure 1. Knowledge Discovery in Database processes
The term Knowledge Discovery in Databases, or KDD
for short, refers to the broad process of finding
knowledge in data. In KDD the most important step is
data mining. KDD will convert the low level data into
high level data. Data mining is filed in which useful
outcome that is being predicted by analyzing large
database. It uses readily built tools to get out the useful
hidden patterns, trends and prediction of future which can
be obtained using the techniques. Data mining involves
various models to discover patterns which consist of
various components. [2]

(IJERCSE)
Vol 6 Issue 1, January 2019
Three steps involved are
• Exploration
• Pattern identification
• Deployment
2. Data Mining Types
Predictive data mining: It produces the model of a system
described by the given data. It uses some variables or
fields in the data set to predict unknown or future values
of other variables of interest.
Descriptive Data Mining: It produces new, significant
information based on the available data set. It focuses on
finding different patterns describing the data that can be
interpreted by human.
3. Data Mining Tasks
• Data processing [descriptive]
• Prediction [predictive]
• Regression [predictive]
• clustering [descriptive]
• Classification [predictive]
• Link analysis associations [descriptive]
4. Prediction on Data Mining Techniques:
There are several data mining techniques which has been
developed and used in data mining projects including
association, classification, clustering, prediction and
neural network.
Classification Analysis Technique:
Classification consists of predicting a certain outcome
based on the given input. This type of technique results
are purely depends on the input and output will be based
on that only. In order to predict the outcome, the
algorithm processes a training set containing a set of
unique attributes and the respective outcome, usually
called destination or prediction attribute. The algorithm
tries to discover correlation between the attributes and to
predict the outcome. Then the algorithm is given a data
set which is not seen before called prediction set, which
contains the same set of attributes, except for the
prediction attribute – not yet known. The algorithm
analyses the input and produces a prediction. The
accuracy of the prediction defines the efficiency of the
algorithm used, for example in a medical database the
training set will have relevant patient information
previously recorded and the prediction attribute has to be
whether the patient had a heart problem or not. Table 1
below illustrates the training and prediction sets of such
database.
Training set- Table 1
Age Heart rate Blood
pressure
Heart
problem
65 78 150/70 yes
37 83 112/76 no
71 67 108/65 no
Prediction set
Age Heart rate Blood
pressure
Heart
problem
43 98 147/89 ?
65 58 106/63 ?
84 77 150/65 ?
Table 1 – Training and Prediction Sets for Medical
Database.
In several types of knowledge representation present in
the literature, classification normally uses prediction
rules to express knowledge. Prediction rules are mostly
expressed in the form of IF-THEN rules, where the
antecedent (IF part) consists of a conjunction of
conditions and the rule consequent (THEN part) predicts
a certain prediction attribute value for an item that
satisfies the antecedent. [4] Using the example above, a
rule predicting the first row in the training set may be
represented as following:
IF (Age=65 AND Heart rate>70) OR (Age>60 AND
Blood pressure>140/70) THEN Heart problem=yes
In most of the cases prediction rule is extremely larger
than the example above. Conjunction has a nice property
for classification; each condition separated by OR’s
defines smaller rules that captures relationship between
attributes. Satisfying any of these smaller rules means
that the effect is the prediction. Each smaller rule is
formed with AND’s which narrow down relationships
between attributes. How predictions are done is measured
in percentage of predictions hit against the total number
of predictions. A probable rule ought to have a hit rate
greater than the occurrence of the prediction attribute. In
other words, if an algorithm is trying to predict rain in a
rainy season and it rains 80% of the time, the algorithm
could easily have a hit rate of 80% by just predicting rain
all the time. Therefore, 80% is the base prediction rate
that any algorithm should achieve in this case. The
perfect solution is a rule with 100% prediction hit rate,
which is very hard, when impossible, to achieve.[4]
Association Rule Learning Technique
An association rule represents a very promising
technique to improve heart disease prediction in medical
industry. Unfortunately, when association rules are

(IJERCSE)
applied on a medical data set, they produce an extremely
large number of rules. Many such rules are medically
irrelevant and the time required to find them can be
impractical. Four constraints were proposed to reduce the
number of rules: item filtering, attribute grouping,
maximum item set size, and antecedent/consequent rule
filtering. A more important issue in general is association
rules are mined on the entire data set without validation
on an independent sample. To solve these, the author has
introduced an algorithm that utilizes search constraints to
reduce the number of rules, searches for association rules
on a training set, and finally validates them on an
independent test set. Instead of using only support and
confidence, one more parameters, that is lift have been
used as the metrics to evaluate the medical significance
and reliability of association rules. Doctors
predominantly use to explicit results. Sensitivity is
defined as the probability of identifying sick patients
accurately, whereas specificity is defined as the
probability of identifying healthy individuals accurately.
Along with confidence Lift is also used to understand
sensitivity and specificity. An algorithm has three major
steps to find predictive association rules in a medical data
set. First, a medical data set with categorical and numeric
attributes is transformed into a transaction data set.
Second, four constraints mentioned above are
incorporated into the search process to find predictive
association rules with medically relevant attribute
combinations in the data. Third, train and test approach is
used to validate association rules.
Types of Association Rule
• Multilevel association rule
• Multidimensional association rule
• Quantitative association rule
• Fuzzy Association Rules Mining
Decision Tree algorithms
It includes CART (Classification and Regression Tree),
ID3 (Iterative Dichotomized 3) and C4.5. These
algorithms differ in selection of splits, when to stop a
node from splitting, and assignment of class to a non-split
node. CART mainly uses Gini index to measure the
impurity of a partition or set of training data computing
or tuples. It can handle high dimension categorical data.
Decision Trees can also handle continuous data (as in
regression) but they must be converted to categorical
data. The decision tree is built from the very small
training set. We will refer to a row as a data instance. The
data set contains three predictor attributes, namely Age,
Gender, Intensity of symptoms and one goal attribute,
namely disease whose values (to be predicted from
symptoms) indicates whether the patient have the disease
or not.
Neural Network
Neural networks got modeled after the cognitive
processes of the brain. They are capable of predicting
new observations from existing. This network consists of
interconnected processing elements also called units,
nodes, or neurons. The neurons within the network work
together in parallel to produce an output function. Since
the computation is performed by the collective neurons, a
neural network can still produce the output function even
if some of the individual neurons are malfunctioning. In
general, each neuron within a neural network has an
associated activation number. Each connection between
neurons has a weight associated with it. These quantities
simulate their counterparts in the biological brain firing
rate of a neuron and strength of a synapse a junction
between two nerves. The activation of a neuron depends
on the activation of the other neurons and the weight of
the edges that are connected to it. The neurons within a
neural network are usually arranged in layers. The
number of layers within the neural network and the
number of neurons within each layer normally matches
the nature of the investigated phenomenon. After the size
has been determined the network is usually then
subjected to training. Then the network receives a sample
training input with its associated classes. Then it applies
an iterative process on the input in order to adjust the
weights of the network so that the future predictions are
optimal. After the training phase, the network is ready to
perform predictions in new sets of data. Neural networks
can often produce very accurate predictions. However,
one of their greatest criticisms is that they represent a
“black-box” approach to the research. They do not
provide any insight into the underlying nature of the
scenario.
Clustering Analysis
Clustering is a process of grouping physical or abstract
objects into classes of similar objects or unsupervised
classification. Clustering analysis helps majorly to
construct meaningful partitioning of a large set of objects
based on a “divide and conquer” methodology which
decomposes a large scale system into smaller
components to simplify the design and implementation.
Clustering is identifying similar groups from an
unstructured data. Clustering is the task of grouping a set
of objects in such a way that object in same group are
more similar to each other to those in other group. Once
the clusters are decided the objects are labeled with their
corresponding clusters, and common features of the
objects in cluster are summarized to form a class
description. For example, a hospital may cluster its
patients in to several groups based on the similarities of

(IJERCSE)
their name, age, sex, disease, symptoms, blood pressure,
sugar level etc., and the common characteristics of the
patients in a group can be used to describe that group of
patients. The doctor has to understand the patients to
predict their disease and provide medication.
Types of clustering methods
• Hierarchical Methods,
• Partitioning Methods,
• Density Based Methods,
• Grid-Based Methods and
• Model Based Algorithms
Categorization of Clustering Algorithms
Algorithms are the key step for solving multiple
techniques. In these clustering techniques, various
algorithms are currently live and still lot more are
evolving, but in general the algorithm for clustering is
neither straight nor canonical.
Hierarchical methods:
• Agglomerative Algorithms and
• Divisive Algorithms.
Partitioning methods:
• Relocation Algorithms,
• Probabilistic Clustering,
• K-Medoids Methods and
• K-Means Methods.
Density-based algorithms:
• Density-based connectivity clustering and
• Density functions clustering.
Grid-based methods:
• Methods based on co-occurrence of categorical data,
• Constraint-based clustering,
• Clustering algorithms used in machine learning,
• Gradient descent and artificial neural networks and
• Evolutionary methods.
Model Based Algorithms:
• Algorithms for high dimensional data,
• Subspace Clustering,
• Projection Techniques and
• Co-Clustering Techniques.
Prediction Techniques
Regression technique shall be adapted for prediction.
Regression analysis shall be used to model the
relationship between one or more independent variables
and dependent variables as well. In data mining
independent variables are attributes already known and
the response variables are what we want to predict.
Unfortunately, many real scenario problems are not
simply prediction. For instance, sales volumes, stock
prices, petrol prices, crude oil prices, gold rates and
product failure rates are all very difficult to predict
because they are very dynamic and may depend on
complex interactions of multiple predictor variables.
Therefore, decision trees, or neural nets are required to
forecast future values. The same model types can often
be used for both regression and classification. For
example, the CART (Classification and Regression
Trees) decision tree algorithm can be used to build both
classification trees (To classify categorical response
variables) and regression trees (To forecast continuous
response variables). Neural networks also can create both
classification and regression models.
TOOLS FOR DATA MINING TECHNIQUES
There are many open source tools available for data
mining. Some of the tools work for clustering, some for
classification, regression, association and some for all.
There are various algorithms for each technique as
discussed above. This section describes features of
different tools and which tool can be used to implement
which algorithm.
1. Rapid Miner
Rapid Miner is a data science software platform that
provides an integrated environment for data preparation,
machine learning, deep learning, text mining and
predictive analysis. It is one of the apex leading open
source systems for data mining.
2. Oracle Data Mining
It represents Oracle’s Advanced Analytics Database.
Leading companies in the Market use it to maximize the
potential of their data to get accurate predictions.
3. IBM SPSS Modeler
It helps to generate data mining algorithms with
minimal or without programming. It is widely used in
anomaly detection, Bayesian networks, CARMA, Cox
regression and basic neural networks use multilayer
perceptron with back-propagation learning.
4. KNIME
Konstanz Information Miner is an open source data
analysis platform. It is a platform that helps to make
predictive intelligence accessible to users without
experience also.
5. Python
It is available as a free and open source language;
commonly business-use case-data visualizations are
straightforward as long as you are comfortable with basic
programming concepts like variables, data types,

(IJERCSE)
6. Orange
It is an open source data visualization, machine learning
and data mining toolkit. It is a component based visual
programming software package for data visualization,
machine learning, data mining and data analysis.
7. Kaggle
It is the world’s largest community of data scientist and
machine learners. Kaggle is a platform that helps to solve
the most difficult problems, recruit strong teams and
accentuate the power of data science in the Industry.
8. Rattle
Rattle GUI is an open and free software package which
provides a graphical user interface for data mining using
R Statistical programming language provided by Toga
ware.
9. Weka
Weka is a platform which supports several standard data
mining tasks, more specifically data pre-processing,
clustering, classification, regression, visualization, and
feature selection. This is developed in Java and the
algorithms can either be applied directly to a dataset or
called from Java code, because of the Java API.
10. Teradata
Teradata analytical platform delivers the best functions
and leading engines to enable users to leverage their
options to choose the tools and languages at scale across
different data types.
Comparative Analysis of Data Mining Tools:
In this section, the best available data mining tools [7]
were taken and comparative study was made by
considering parameter license, technical specification,
features and describing in table.
Table2
Features of the most commonly used data mining tools.
Software tool Types Features
WEKA Machine Learning based It is java based open source data mining tool,
based on various data mining and machine
learning algorithms.
ORANGE Machine Learning, Data mining, data
visualization
It is open source data mining tool, based on
Visual programming, visualization, large tool
box, plate form independent, and interaction and
data analysis.
Rapid Miner Statistical analysis, predictive
analysis’
It is an important predictive analytic platform. It
is user friendly, rich library of data science and
has many machine learning algorithms.
Data Melt Statistical analysis , numeric and
symbolic computations, scientific
visualization
It is based on cluster analysis, linear regression,
neural networks, curve fitting, fuzzy system,
analytic calculations & interactive visualizations.
Apache Mahout Machine Learning based It is a library of machine learning algorithms
which is used in clustering, classification &
frequent pattern data mining. It is also used in a
distributed mode with integration of Hadoop
ELKI cluster analysis and outlier detection It is java based open source data mining tool and
licensed under AGPLv3.It focuses on outlier
detection and cluster analysis with a compilation
of several algorithms from both these domains.
KNIME Enterprise reporting, Business
Intelligence
It is based on java and built upon Eclipse. It is
scalable, highly extensive, and capable of data
visualization and import –export workflow.
MOA Machine Learning based It can handle large volumes of real time data
streams at a very high speed and can be used
through GUI, command line, or Java API.

(IJERCSE)
KEEL Machine Learning based It is java based tool and licensed under
GPLV3and based on clustering, classification,
and association. It has a very user friendly GUI
Rattle or R Statistical computing It is a statistical programming language and R
program can run on Mac OS Linux, and
Windows, it is based on statistics, clustering,
modeling and visualization.
II. CONCLUSION
Data mining offers an appropriate way to discover hidden
patterns by various prediction analysis within large
amounts of data. These hidden patterns can be potentially
used to predict future behavior. The availability of new
data mining algorithms is relatively more however; it
should be met with caution. There have been a large
number of data mining algorithms rooted in this field to
perform different data analysis tasks and take multiple
decisions across various decision making panel and
different industries. This paper dealt with a detailed study
on prediction of Data Mining, its techniques, tasks and
related tools. Data mining has wide application domain
almost in every industry where the data is generated
that’s why data mining is considered as one of the most
important frontiers in database and information systems,
data base management systems and one of the most
promising interdisciplinary developments in Information
Technology. Data mining helps researchers to analyze
past data sets, practices, progress, and nature of
performances and helps them to predict results, forecast
budgets and also to take various constructive decisions in
the industry to get better results towards growth.
REFERENCES
[1] Kumbhare, Trupti A., and Santosh V. Chobe. "An
overview of association rule mining algorithms."
International Journal of Computer Science and
Information Technologies 5.1 (2014): 927-930.
[2] Ramageri, Bharati M., and B. L. Desai. "Role of data
mining in retail sector." International Journal on
Computer Science and Engineering 5.1 (2013): 47.
[3]Ilayaraja, M., and T. Meyyappan. "Mining medical
data to identify frequent diseases using Apriori
algorithm." Pattern Recognition, Informatics and Mobile
Engineering (PRIME), 2013 International Conference on.
IEEE, 2013.
[4] Voznika, Fabricio, and Leonardo Viana. "Data
Mining Classification." (2007).
[5] Tamilselvi, R., and S. Kalaiselvi. "An Overview of
Data Mining Techniques and Applications." Int J Sci
Eng. Res 1.1-3 (2013): 506-9.
[6] Karur Parminder and Qamar Parvez Rana.
“Comparison of various data mining tools”, International
Journal of Engineering Research & Technology
(IJERT)ISSN: 2278-0181IJERTV3IS100246 Vol. 3 Issue
10, October- 2014
[7] Rawat, Keshav Singh, and I. V. Malhan.
“Comparative Analysis of Data Mining Techniques,
Tools and Machine Learning Algorithms for Efficient
Data Analytics." IOSR Journal of Computer Engineering
ISSN: 2278-0661,p-ISSN: 2278-8727, Volume 19, Issue
4,
[8] Doddi, Achla Marathe, SS Ravi, David C. Torney,
Srinivas. "Discovery of association rules in medical
data." Medical informatics and the Internet in medicine
26.1 (2001): 25-33.
[9] Gera, Mansi, and Shivani Goel. "Data mining-
techniques, methods and algorithms: A review on tools

(IJERCSE)
and their validity." International Journal of Computer
Applications 113.18 (2015).
[10] Verma, Manish, et al. "A comparative study of
various clustering algorithms in data mining."
International Journal of Engineering Research and
Applications (IJERA) 2.3 (2012): 1379-1384.

prediction using data mining.pdf

Recommended

More Related Content

Similar to prediction using data mining.pdf (20)

More from NavAhmed3 (7)

Recently uploaded (20)

prediction using data mining.pdf