Recent Survey: Prediction for Heart Attack Problem Using Various Classification Techniques

Recent Survey: Prediction for Heart Attack Problem Using Various Classification Techniques

Prachi Paliwal

Dept. of Computer Science & Engineering(Software Engineering)

J.I.T., Borawan Khargone, India

Mr. Mahesh Malviya

Assistant Professor

Dept. of Computer Science & Engineering Khargone, India

ABSTRACT Data mining methods are used to analyze the medical data contents. Superior data mining techniques are developed and used to discover hidden pattern form historical data. New Models are developed from these techniques will be useful for medical practitioners to take successful decision. Diagnosis of heart attack is a significant task in medical science. The term Heart attack includes the various diseases that involve the heart attack problem. The exposure of heart attack problem from different symptoms is an important issue for predicting heart attack problem. This research paper includes the study of various classification techniques like Decision Tree Induction, Support Vector Machines (SVM) , Bayesian Classification, Rule- based classification, Classification by back propagation, Neural Network as a Classifier The k-Nearest Neighbor Algorithm and Classification using Genetic Algorithms (GA) .

Keywords Data Mining, Diagnosis, Heart Attack, Symptoms, Classification, Prediction.

1. INTRODUCTION

   Classification is a process which classifies data based on the training set and the values in a classifying attribute and uses it in classifying new data. Classification predicts categorical class labels. Classification is divided into two-step.

   Accuracy

   1. Constructing a Model :

      This step describing a set of predetermined classes. Each tuple/sample is assumed to belong to a predefined class, as determined by the class label attribute. The set of tuples used for model construction is training set. The model is represented as classification rules, decision trees, or mathematical formulae.

   2. Usages constructed Model:

   This step classifying future or unknown objects.

   Figure 1.1 represent the system architecture for classification process in which sample data are used to construct model and then tested this model for accuracy to the a given tuple predicted the class level.

2. EVALUATING PARAMETERS

   There are some parameters which are used to evaluate classification methods. These parameters are:

   A. Accuracy :-

   This include accuracy of the classifier in term of predicting the class label,guessing value of predicted attributes. Accuracy can be estimated using one or more test sets that are independent of the training set.

   1. Speed:-

      Disease Diagnosis Data

      Disease Treatment Data

      Measurement

      Data mining Classification techniques

      This include the required time to construct the model (training time) and time to use the model (classification/prediction time). In other word this refers to the computational costs.

   2. Robustness:-

   This is the ability of the classifier or predictor to make correct predictions given noisy data or data with missing values.

   C . Scalability:-

   Efficiency in term of database size.

   Classifying the tuple into predicting class level

   Figure 1.1 System architecture for classification process

   1. Interpretability:-

      Understanding and insight provided by the model. Interpretability is subjective and therefore more difficult to assess.

   2. Other measures:-

      Includes goodness of rules, such as decision tree size or compactness of classification rules.

3. VARIOUS CLASSIFIERS

   1. Decision Tree Classifier

   Decision tree is A flow-chart-like tree structure Leaf nodes represent class labels or class distribution. Decision tree is a classifier in which each non-terminal node represents either a test or decision for the given data item. Which branch to be select next is depends upon the outcome of the test. To classify a given data item, need to from start at the root node and follow the assertions down until we reach a terminal node or leaf node.

   Age

   C) Neural Network as a Classifier

   Neural network approach has been widely adopted as classifiers. The neural network provides several advantages, like arbitrary decision its nonparametric nature, boundary capability, easy adaptation to different types of data. Neural nets may used in classification problems where the output is a categorical variable. Neural nets has three layers first is input layer, seconds is hidden layer and third output layer. The nodes in the input layer linked with a number of nodes in the hidden layer. Each input node joined to each node in the hidden layer. The nodes in the hidden layer may connect to nodes in another hidden layer, or to an output layer. The output layer consists of one or more response variables. There is numerous advantages of ANN some of these include-

   1. Good Accuracy.

   2. Free from prior assumptions about the distribution of the data.

   3. Noise tolerance.

   4. ANN can be implemented in parallel hardware.

   Youth

   Student

   age

   Middle age

   Yes

   Senio r

   Credit rating

   Patient database

   No Yes Fair Excellen t

   Calculate probability of each attribute

   No Yes

   No Yes Naive Bayesian classifier

   Figure 1.2 simple decision tree classification

   Decision is made when a terminal node is approached. Decision trees use recursive data partitioning. The important

   Predict the given tuple

   things in decision tree are attribute selection measure. There

   is important parameter used for attribute selection. The attribute with highest information gain is used to be selected as a root.

   B) Naive Bayesian Classifiers

   The Naive Bayesian classifier or simple Bayesian classifiers statistical classifiers and able to predict class membership probabilities, such as the probability that a given tuple belongs to a particular class. Bayesian classification is based on Bayes theorem. The Naive Bayes Classifier technique is particularly suited when the dimensionality of the inputs is high. Naive Bayesian classifier algorithm is used to create models with predictive capabilities. It provides new ways of exploring and understanding data. Figure 1.3 shows the working of Naive Bayesian classifiers.

   Figure 1.3 simple modelfor Naive Bayesian Classifiers

   D) Using IF-THEN Rules as Classifier

   A rule-based classifier uses a set of IF-THEN rules for classification. An IF-THEN rule is an expression of the form

   IF condition THEN conclusion. An example is rule R1,

   R1: IF age = youth AND student = yes THEN buys computer

   = yes.

   The IF part of a rule is known as the rule antecedent or precondition. The THEN part is the rule consequent. In the rule antecedent, the condition consists of one or more attribute tests (such as age = youth, and student = yes)

   that are logically ANDed. The rules consequent contains a class prediction (in this case, we are predicting whether a customer will buy a computer). R1 can also be written as

   R1: (age = youth) ^ (student = yes)) (buys computer =

   yes).

   If the condition (that is, all of the attribute tests) in a rule antecedent holds true for a given tuple, we say that the rule antecedent is satisfied and that the rule covers the tuple.

4. LITERATURE REVIEW

   In 2010 O.P.V Yas and Sunita Soni proposed Using Associative Classifiers for Preictive Analysis in Health Care Data Mining. They describe that analysis technique to discover a small set of rule in the database to forms an accurate classifier Association rule mining is important. They introduce the combined approach that integrates association rule mining and classification rule mining. This is new classification approach is implemented by focusing on mining a special subset of association rules called classification association rule, then classification is being performed using rules. The associative classifiers are especially fit to applications were the model may assist domain experts in their decisions There are many associative classification approaches that have been proposed recently such as CBA, CMAR, CPAR and MCAR and MMAC.

   In 2011 Mai Shouman, Tim Turner, Rob Stocker proposed

   Using Decision Tree for Diagnosing Heart Disease Patients

   . They show that Decision Tree is one of the successful data mining techniques used in the diagnosis of heart disease. Yet its accuracy is not perfect. The proposed work systematically tested combinations of discretization, decision tree type and voting to identify a more robust, more accurate method. They investigate a range of techniques to different types of Decision Trees seeking better performance in heart disease diagnosis and proposed a model that outperforms.

   In 2012 M.Akhil jabbar , Dr.Priti Chandrab , Dr.B.L Deekshatulu Proposed Heart Disease Prediction System using Associative Classification and Genetic Algorithm. The main advantage of genetic algorithm is the discovery of high level prediction rules is that the discovered rules are highly comprehensible, having high predictive accuracy and of high interestingness values. The proposed method helps in the best prediction of heart disease which even helps doctors in their diagnosis decisions.

   In 2012 Sunita Soni and O. P. Vyas proposed Fuzzy Weighted Associative Classifier A Predictive Technique for Health Care Data Mining. They extend classification problem using Fuzzy Association Rule Mining and proposed the concept of Fuzzy Weighted Associative Classifier. Domain experts like models are fir for Associative classifiers in their decisions. They proposed a new Fuzzy Weighted Associative Classifier that generates classification rules using Fuzzy Weighted Support and Confidence framework. They proposed a theoretical model to introduce new associative

   classifier that takes advantage of Fuzzy Weighted Association rule mining.

   In 2012 Sulabha S. Apte, Ph.D. and Chaitrali S. Dangare proposed Improved Study of Heart Disease Prediction System using Data Mining Classification Techniques. The proposed work has analyses prediction systems for Heart disease using more number of input attributes. The system uses medical terms as an attributes to predict the likelihood of patient getting a Heart disease. Total 13 attributes are used for prediction. Two more attribute obesity and smoking added. The data mining classification techniques, namely Decision Trees, Naive Bayes, and Neural Networks are analyzed on Heart disease database. The performance of these techniques is compared, based on accuracy.

   In 2013 M. Akhil Jabbar, B.L Deekshatulu and Priti Chandra proposed Classification of Heart Disease using Artificial Neural Network and Feature Subset Selection. They have proposed a new feature selection method using ANN for heart disease classification. For rank the attributes which contribute more towards classification of heart disease they applied different feature selection methods, and indirectly reduce the no. of diagnosis tests to be taken by a patient. The proposed method eliminates useless and distortive data. The proposed method will contribute reliable and faster automatic heart disease diagnosis system, where easy diagnosis of heart disease will saves lives.

   In 2013 V.V.Jaya Rama krishniah, D.V.Chandra Sekar and Dr.K.Ramchand H Rao proposed Predicting the Heart Attack Symptoms using Biomedical Data Mining Techniques. This proposed based on clustering based k- Nearest Neighbor, K Mean and Entropy based mean clustering algorithms. The performance comparison made on Entropy based mean the best compact time for processing dataset. The proposed work shows the enhanced performance according to the attribute.

   [7]In 2014 N. S. Nithya and K. Duraiswamy proposed Gain ratio based fuzzy weighted association rule mining classifier for medical diagnostic interface. Earlier model based on information gain and fuzzy association rule mining algorithm for extracting both association rules and membership functions are not feasible.

   When taking a large number of distinct values. So they modify gain ratio based fuzzy weighted association rule mining and improve the classifier accuracy.

   Decision Tree
   Benefits	Limitations
   No requirements of domain	1.It generates categorical output
   knowledge in the construction	2. Classifier is depend upon the
   of decision tree.	type of dataset
   High dimension data can easily	3. It is restricted to one output
   process.	attribute
   Decision tree assigns exact
   values to outcomes

   BENEFITS AND LIMITATIONS

   Bayesian Classification
   Benefits	Limitations
   1. Naive Bayesian	Naive Bayesian classifiers is
   classifiers makes	a probability based methods.
   computational process	It does not give accurate
   easy.	results
   2. Naive Bayesian	2. class conditional
   classifiers provides better	independence, therefore loss
   speed and accuracy for	of accuracy
   huge datasets

5. CONCLUSION

There are various classification techniques that can be used for the identification and prevention of heart disease. The performance of classification techniques depends on the type of dataset that we have taken for doing experiment. Classification techniques provide benefit to all the people such as doctor, healthcare insurers, patients and organizations who are engaged in healthcare industry. Decision tree, Bays Naive classification, Support Vector Machine, Rule based classification, Neural Network as a classifier etc. These techniques are compared on basis of Sensitivity, Specificity, Accuracy, Error Rate, True Positive Rate and False Positive Rate. The objective of each techniques is to predict more accurately the presence of heart disease with reduced number of attributes.

REFERENCE

1. Divya Tomar and Sonali Agarwal A survey on Data Mining approaches for Healthcare International Journal of Bio-Science and Bio-Technology Vol.5, No.5 (2013), pp. 241-266.

2. Bangaru Veera Balaji and Vedula Venkateswara Rao Improved Classification Based Association Rule Mining International Journal of Advanced Research in Computer and Communication Engineering Vol. 2, Issue 5, May 2013

3. V. Krishnaiah, Dr. G. Narsimha and Dr. N. Subhash Chandra Diagnosis of Lung Cancer Prediction System Using Data Mining Classification Techniques (IJCSIT) International Journal of Computer Science and Information Technologies, Vol. 4 (1) 2013, 39 45

4. Shamsher Bahadur Patel, Pramod Kumar Yadav and Dr. D. P.Shukla

   Predict the Diagnosis of Heart Disease Patients Using Classification Mining Techniques IOSR Journal of Agriculture and Veterinary Science (IOSR-JAVS) e-ISSN: 2319-2380, p-ISSN: 2319-2372. Volume 4, Issue 2 (Jul. Aug. 2013),

5. N. Suneetha ,Ch.V.M.K.Hari and Sunil Kumar Modified Gini Index Classification: A Case Study Of Heart Disease Dataset (IJCSE) International Journal on Computer Science and Engineering Vol. 02, No. 06, 2010, 1959-1965

6. Jyoti Soni, Uzma Ansari, Dipesh Sharma and Sunita Soni

   Intelligent and Effective Heart Disease Prediction System using Weighted Associative Classifiers Jyoti Soni et al. / International

   Journal on Computer Science and Engineering (IJCSE) ISSN : 0975- 3397 Vol. 3 No. 6 June 2011

   Neural Network
   Benefits	Limitations
   1. Able to handle noisy	1. Neural Network has the
   data.	problem Over-fitting
   2. Easily identify complex	2. Neural Network has the
   relationships	problem Local minima

7. Sunita Soni and O.P.Vyas Using Associative Classifiers for Predictive Analysis in Health Care Data Mining International Journal of Computer Applications (0975 8887) Volume 4 No.5,

   July 2010

8. Mai Shouman, Tim Turner, Rob Stocker Using Decision Tree for Diagnosing Heart Disease Patients Proceedings of the 9-th Australasian Data Mining Conference (AusDM'11), Ballarat,

   Australia

9. M. Akhil jabbar, Dr B.L Deekshatulu and Dr Priti Chandra Heart Disease Classification Using Nearest Neighbor Classifier With Feature Subset Selection Computer Science and Telecommunications 2013|No.3(39)

10. Sunita Soni and O.P.Vyas Fuzzy Weighted Associative Classifier: A Predictive Technique For Health Care Data Mining International Journal of Computer Science, Engineering and Information Technology (IJCSEIT), Vol.2, No.1, February 2012

11. Chaitrali S. Dangare and Sulabha S. Apte, PhD. Improved Study of Heart Disease Prediction System using Data Mining Classification Techniques International Journal of Computer Applications (0975 888) Volume 47 No.10, June 2012

12. M. Akhil Jabbar, B.L Deekshatulu & Priti Chandra Classification of Heart Disease using Artificial Neural Network and Feature Subset Selection Global Journal of Computer Science and Technology Neural & Artificial Intelligence Volume 13 Issue 3 Version 1.0 Year 2013 Type: Double Blind Peer Reviewed International Research Journal

    Publisher: Global Journals Inc. (USA) Online ISSN: 0975-4172 & Print ISSN: 0975-4350

13. N S Nithya and K Duraiswamy Gain ratio based fuzzy weighted association rule mining classifier for medical diagnostic interface Vol. 39, Part 1, February 2014, pp. 3952. Indian Academy of Sciences

14. M.Akhil jabbar, Dr.Priti Chandrab , Dr.B.L Deekshatuluc Heart Disease Prediction System using Associative Classification and Genetic Algorithm International Conference on Emerging Trends in Electrical, Electronics and Communication Technologies-

    ICECIT, 2012

15. A. Anushya and A. Pethalakshmi A Comparative Study of Fuzzy Classifiers With Genetic On Heart Data International Conference on Advancement in Engineering Studies & Technology, ISBN : 978- 93-81693-72-8, 15th JULY, 2012, Puducherry