Numerous work has
been done related to heart prediction system by using various data mining
techniques and algorithms by many authors. The aim of all is to achieve better
accuracy and to make the system more efficient so that it can predict the
chances of heart attack. This paper aims at analyzing the various data mining
techniques introduced in recent years for heart disease prediction. Different
data mining techniques have been used in the diagnosis over different Heart
disease datasets. Knowledge of the risk factors associated with heart disease
helps health care professionals to identify patients at high risk of having
heart disease. Statistical analysis has identified the risk factors associated
with heart disease to be age, blood pressure, smoking habit, total cholesterol,
diabetes, hypertension, family history of heart disease, obesity, and lack of physical
activity.

            In paper 1, it describes about heart
disease prediction were further analysed using three data mining classification
techniques namely decision tree, artificial neural network, and SVM. The
results were compared and the accuracy obtained were as follows: 79.05%,
80.06%, and 84.12%, respectively. Their analysis shows that out of these four
classification models, SVM predicts heart disease with the highest accuracy.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!


order now

            In paper2, it describes the
approaches to identify the risk factors from the extracted itemsets that cause
heart disease. Here it surveys various latest frequent pattern mining
algorithms on data streams to understand various advantages and disadvantages,
so they provides a way of using new insights in the direction of frequent
pattern.

            In
paper3, it describes that the Neural Networks with 15 attributes has
outperformed over all other data mining techniques. Decision Tree has shown
good accuracy with C4.5, ID3, CART and J48. Decision Tree has shown good
accuracy with the help of genetic algorithm and feature subset selection. Naïve
Bayes algorithm gives an average prediction with 90% accuracy. The following
table shows the interpretation of various research papers we have studied

//In
paper4, it describes BPNN and
BNN gave the highest classification accuracy of 78.43 %, while RBF kernel SVM
gave the lowest classification accuracy of 60.78 %. BNN presented the best
sensitivity of 96.55 % and RBF kernel SVM displayed the lowest sensitivity of
41.38 %. Both polynomial kernel SVM and RBF kernel SVM presented the minimum
and maximum specificity of 45.45 % and 86.36 %, respectively.

In
paper5, it describes that fast algorithms such as decision tree have relatively
poor accuracy compared to other knowledge models like neural networks. In order
to overcome this problem, a large number of decision trees are generated for
the same data set, and used simultaneously for prediction. Random forest is one
such ensemble based method which is commonly used with decision trees. This
System mainly focuses on the supervised learning technique called the Random
forests for classification of data by changing the values of different hyper
parameters in Random Forests Classifier to get accurate classification results.

 

 In paper 6, it describes about three
Classification function Techniques in Data mining are compared for predicting
Heart Disease with reduced number of attributes .They are Naïve Bayes, Decision
Tree and Classification by Clustering. Here Genetic algorithm is used to
determine the attributes which contribute more towards the diagnosis of heart
ailments which indirectly reduces the number of tests which are needed to be
taken by a patient. Fourteen attributes are reduced to 6 attributes using
genetic search. The observations exhibit that the Decision Tree data mining
technique outperforms other two data mining techniques after incorporating
feature subset selection with relatively high model construction time. Naïve
Bayes performs consistently before and after reduction of attributes with the
same model construction time. Classification via clustering performs poor
compared to other two methods.

In paper 7, two experiments were conducted with
all 13 attributes and with 6 attributes of reduced dataset by applying
attribute selection method. The observation was that SVM (97.9%, 89.4%), Simple
logistic (69.2%, 71.6%) and Multilayer perceptron (74.3%, 79.1%) techniques are
achieved different accuracy in two scenario. From this it shows that SVM has
gretest accuracy.