ABSTRACT
The early prognosis of cardiovascular diseases can aid The heart disease database is pre-processed
pre to make
in making decisions to lifestyle changes in high risk the mining process
rocess more efficient. The pre-processed
patients and in turn reduce their complications. data is classified with Regression.
Research has attempted to pinpoint the most
influential factors of heart disease as well as LITERATURE SURVEY
accurately
tely predict the overall risk using homogenous
data mining techniques. Recent research has delved Carlos Ordonez [14] did a study on prediction of heart
into amalgamating these techniques using approaches disease with the help of Association rules. They used
such as hybrid data mining algorithms. This paper a simple mapping algorithm. This algorithm
proposes a rule based model to compare the constantly treats attributes as numerical or categorical.
accuracies of applying rules to the individual results This is used to convert medical records to a
of logistic regression on the Cleveland Heart Disease transaction
ransaction format. An improved algorithm is used to
Database in order to present an accurate model of mine the constrained association rules. A mapping
predicting heart disease. table is prepared and attribute values are mapped to
items. The decision tree is used for mining data
KEYWORDS: heart disease prediction, logistic because they automatically Split numerical values
regression, Cleveland heart disease data base [14].
4]. The split point chosen by the Decision tree is of
little use only. Clustering is used to get a global
INTRODUCTION understanding of data.
This paper analyzes the heart disease predictions
Usha Rani [15] have proposed a system for predicting
using classification algorithms. These
ese hidden patterns
heart disease with the help of artificial neural
can be used for health diagnosis in Medicinal data.
network, which is a combination
ination of feed forward and
Data mining technology afford an effective approach
back propagation algorithm. The experiment is carried
to latest and indefinite patterns in the data. The
out by considering single and multilayered neural
information which is identified can be used by the
network models. Parallelism is implemented to speed
healthcare administrators to get better services. Heart
up the learning process at each neuron in all hidden
disease was the most important reason of victims in
and output layers.
the countries like India, United States. Data mining
techniques like Association Rule Mining, Clustering,
T. Revathi and S. Jeevitha [16] analyzed the data
Classification algorithms such as Decision tree, C4.5
mining algorithms on prediction of heart disease. The
algorithm.
clinical data related to heart disease is used for
The dataset consists of 15 types of attributeslisted in So, the main motto of Logistic regression is to
the table 1 determine the result of each variable correctly
Logistic regression is also known as logistic model/
logit model that provide categorical variable for target
variable with two categories such as light or dark,
slim/ healthy.
TECHNIQUES USED
REGRESSION
@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 2 | Issue – 3 | Mar-Apr 2018 Page: 1468
International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
2456
FLOW CHART CONCLUSION
In conclusion, as identified through the literature
review, there is a need for combinational and more
complex models to increase the accuracy of predicting
the early onset of cardiovascular diseases. This paper
proposes a framework using combinations of support
vector machines, logistic regression, and decision
trees to arrive at an accurate prediction of heart
disease. Using the Cleveland Heart Disease database,
this paper provides guidelines to train and test the
system and thus attain
tain the most efficient model of the
multiple rule based combinations. Further, this paper
proposes a comparative study of the multiple results,
which include sensitivity, specificity, and accuracy. In
addition, the most effective and most weighed model
cann be found. Further work involves development of
the system using the mentioned methodologies and
thus training and testing the system. Future work may
also involve the development of a tool to predict the
risk of disease of a prospective patient. The
framework
ework can also be extended for use on other
models such as neural networks, ensemble algorithms,
flow chart diagrams used for our study etc.
REFERENCES
RESULT
1) Mackay,J., Mensah,G. 2004 “Atlas of Heart
Disease and Stroke” Nonserial Publication, ISBN-
ISBN
139789241562768 ISBN-1010 9241562765.
2) Robert Detrano 1989 “Cleveland Heart Disease
Database” V.A. Medical Center, Long Beach and
Cleveland Clinic Foundation.
3) Yanwei Xing, Jie Wang and Zhihong Zhao
Yonghong Gao 2007 “Combination data mining
methods with new medical data to predicting
outcome of Coronary Heart
Hear Disease”
Convergence Information Technology, 2007.
International Conference November 2007, pp 868-
868
872.
4) Jianxin Chen, Guangcheng Xi, Yanwei Xing, Jing
Chen, and Jie Wang 2007 “Predicting Syndrome
by NEI Specifications: A Comparison of Five
Data Mining Algorithms
orithms in Coronary Heart
Disease” Life System Modeling and Simulation
Lecture Notes in Computer Science, pp 129-135.
129
In this way the heart disease is predicted accurately 5) Jyoti Soni, Ujma Ansari, Dipesh Sharma 2011
and easily by using the logistic regression and above “Predictive Data Mining for Medical Diagnosis:
flowchart’s.
’s. Result of the study contains 2 variables An Overview of Heart Disease Prediction”
one is detected and other is not detected. International
ational Journal of Computer Applications,
doi 10.5120/2237- 2860.
@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 2 | Issue – 3 | Mar-Apr 2018 Page: 1470