Significant Features and Identification of Heart disease

prediction by using data mining techniques

Authors Name/s per 1st Affiliation (Author) Authors Name/s per 2nd Affiliation (Author)
line 1 (of Affiliation): dept. name of organization line 1 (of Affiliation): dept. name of organization
line 2-name of organization, acronyms acceptable line 2-name of organization, acronyms acceptable
line 3-City, Country line 3-City, Country
line 4-e-mail address if desired line 4-e-mail address if desired

Abstract— In day to day life, there are various factors that similar as rotundity, family history, poor diet, high blood
affect the mortal heart. Numerous problems are being at a pressure, idle geste, high blood cholesterol, family history,
rapid-fire pace and novel heart conditions are fleetly smoking and hypertension. The opinion of heart conditions
identified. In this stressful of world, Heart, being an is an important and is the most complicated task in medical
essential organ in the body pumps blood through the body field.
for blood rotation essential and its health is to be conserved
for a healthy living. The main provocation of doing this All the mentioned factors are taken into notable
design is to provide a heart complaint prediction model for consideration when assaying as well as understanding the
the prediction of circumstances of heart complaints. Further, cases by croaker through homemade check-ups at regular
this exploration study is aimed towards relating the intervals of time. The heart complaint symptoms greatly
algorithms to relate the possibility of heart complaint in a depend upon which of discomfort felt by an existent. Some
case. The identification of possibility of heart complaints in symptoms are not generally linked by the common people.
a person is complicated process for medical interpreters
because it takes times of experience and violent medical Still, common symptoms include chest pain,
tests need to be conducted. In this study, two data mining breathlessness, and heart pulsations. The chest pain
algorithms such as KNN and SVM classification are common to numerous types of heart complaint is known as
addressed and used to develop the prediction system in angina, or angina pectoris, and occurs when a part of heart
order to dissect as well as prognosticate the possibility of doesn't admit sufficient oxygen. Angina is started by
heart complaint. The main idea of the significant stressful events/physical exertion and typically lasts under
exploration work is to identify algorithms suitable to ten twinkles.
provide maximum accuracy when classification of
normal/abnormal person is carried out. Therefore prevention Heart attacks also do as a result of various types of
of loss of lives at an earlier stage is now possible. It is sure heart complaint. The sign of the heart attack is analogous to
that the above algorithms perform better when compared to angina except that they do during rest and tend to be much
other algorithms for heart complaint prediction. The design more severe. The symptoms of the heart attack can
is designed using Python 3.7. occasionally act indigestion.

Keywords— Data mining, Prediction model Classification Heartburn and a stomach pang do, as well as a heavy
algorithms, Feature selection, Heart disease prediction feeling in the chest. Other symptoms of heart attack include
pain travels through the body, for illustration from casket to
I. INTRODUCTION the arms, neck, back, tummy, or jaw, flightiness and dizzy
There may also be several inheritable factors through sensations, nausea and vomiting and gushing sweating,.
which a heart complaint type is passed down from Heart failure is an outgrowth of heart complaint, and
generations. According to World Health Organization, every breathlessness do when the heart becomes too weak to
time more than twelve million deaths are being worldwide blood circulation.
due to colorful types of heart conditions which is also
known by cardiovascular complaint. Some heart conditions do with no symptoms at each,
especially in aged grown-ups and individualities with
The term heart complaint includes various conditions diabetes. The term' natural heart complaint'covers a range of
that are different and specifically affect heart and highways conditions, but the general symptoms include sweating, high
of d:a mortal being. Indeed youthful aged people around situations of fatigue, fast twinkle and breathing,
their 20-30 times of lifetime are getting affected by heart breathlessness, casket pain. Still, these symptoms might not
conditions. develop until a person is aged than 13 times.

The increase in possibility of heart complaint among In these types of cases, the opinion becomes an
youthful may be due to bad eating habits, restless nature, intricate task taking great experience and high skill. A threat
lack of sleep, depression and multitudinous other factors of a heart attack or the possibility of the heart complaint if
linked beforehand, can help the cases take preventives and They performed classical i) statistical analysis and
take nonsupervisory measures. Lately, the healthcare ii) data mining analysis using substantially Bayesian
assiduity has been generating huge quantities of data about networks. The mean age of 533 cases was 63 (± 17) and
cases and their complaint opinion reports are being sample was composed of 390 (72) men and 143 (28)
especially taken for the analysis of heart attacks worldwide. women. Cardiac arrest was observed at home for 412 (77)
When the data about heart complaint is huge, the machine cases, in public place for 62 (12) cases and on public trace
literacy ways can be enforced for the analysis. for 60 (11) cases. The belief network of variables showed
that the remaining alive probability after heart failure is
Data Mining is a task of rooting the vital decision directly associated to 5 variables coitus, age, the original
making information from a philanthropy of once records for cardiac meter, the origin of the heart failure and technical
unborn analysis or vaticination. The information may be reanimation ways employed.
hidden and isn't identifiable without the use of data mining.
Data booby-trapping styles help clinicians to
The classification is one data mining fashion through prognosticate the survival of cases and also acclimatized
which the unborn outgrowth or prognostications can be their practices. This work was carried out for each medical
made grounded on the literal data that's available. The procedure and medical problem and it came possible to
medical data mining made a possible result to integrate the make a decision tree fleetly with data of a service. The
bracket ways and give motorized training on the dataset that comparison among classic analysis and data mining analysis
further leads to exploring the retired patterns in the medical showed us the donation of data mining system for sorting
data sets which is used for the analysis of the case’s unborn variables and conclude on significance or the impact of data
state. Therefore, by using medical data booby-trapping it's and variables on the criterion of study. The main limit of
possible to give perceptivity on a case’s history and is system is knowledge accession and necessity to obtain
suitable to give clinical support through the analysis. For sufficient data to yield an applicable model.
clinical analysis of the cases, these patterns are veritably
important essential. In simple English, the medical data Cardiac arrest is denoted as a robotic unrecoverable
mining uses bracket algorithms that are a vital part for arrest of the general rotation by the cardiac inefficacity. It's
relating the possibility of heart attack before the honored with the absence of femoral palpitation for further
circumstance. than 5 seconds. Without reanimation, the cardiac arrest
leads to the unforeseen cardiac death. The public health
The classification algorithms can be trained and tested impact of the unforeseen cardiac death is heavy since
to make the prognostications that determine the person’s survival rate is estimated at between one and twenty for
nature of being affected by heart complaint. In this cardiac arrest cases. This represents to deaths the time in
exploration work, the supervised machine learning United States and to deaths in France. The profiles of the
conception is employed for making the prognostications. A cases are now well known since it generally concern men
relative analysis of the three data mining bracket algorithms from about 40 to 75 times.
videlicet Random Forest, Decision Tree and Naïve Bayes
are used to make prognostications. The analysis is done at Hospitalization should be optimal and fast.
several situations of cross confirmation and several chance According to type of cardiac attacks, procedure of
of chance split evaluation styles independently. supposition of the responsibility varies and some studies
show interest of ways over others, according to the cause of
The StatLog dataset from UCI machine learning the cardiac arrest. Heart complaint is number one cause of
depository is employed for making heart complaint death in the U.S. According to the American Heart
prognostications in this exploration work. The Association, an estimated people in the U.S. have a
prognostications are made using the bracket model that's coronary attack each time.
erected from the bracket algorithms when the heart
complaint dataset is used for training. This final model can Ninety five percent of unforeseen cardiac arrest
be used for analysis of any types of heart conditions. patients die before reaching sanitarium and heart complaint
further lives each time than following six leading causes of
II. RELATED WORKS death combined (habitual lower respiratory conditions,
cancer, accidents, influenza, diabetes mellitus, and
pneumonia). Nearly people in the U.S. die from heart
In this paper (1) the authors stated that prognostic complaint each time are below the age of 65. These data
of life for cases with heart failure remains poor. By using show that the interest to prognosticate the death threat after
data mining techniques, the purpose of this study is to heart failure and the requirement to dissect events that
estimate the most important criteria to prognosticate patient passed during care to give the prognostic information.
survival and to outline cases for estimating their survival Classic statistical analyses have formerly been
chances together with the most suitable fashion for health done and gave some information related with epidemiology
care. Five hundred and thirty three cases are suffered from of the heart failure and failure causes. This paper presented
cardiac arrest included in the analysis. the use of the probability in statistical approach to outline
heart failure in case samples and prognosticate the impact of a growing consummation/anticipation that data, intelligently
events in the care process. anatomized and are presented, will be the precious resource
to be used for the competitive advantage. The computer
They concluded that it seems that the use of wisdom community is responding to the scientific and
Bayesian network in medical analysis is useful to explore practical challenges presented by need to get the knowledge
data and to find hidden connections between events and or adrift in the deluge of data.
characteristics of the samples. It is the first approach to
agitate suppositions between clinicians and statistical In assessing the eventuality of AI technologies,
experts. The main limit of the tools is the necessity to have Michie (1990), a leading European expert on machine
sufficient data to find chronicity in the connections. literacy, prognosticated that “ the coming area that's going
to explode is the use of machine literacy tools as a element
In this paper (2) the authors stated that after a of large-scale data analysis.” A recent National Science
decade of abecedarian interdisciplinary exploration in the Foundation factory on the future of database exploration
machine literacy, spadework in this field has been done; ranked data mining among the most promising exploration
1990s should see wide exploitation of the knowledge motifs for the 1990s.
discovery as an aid to assembling knowledge bases. The
contributors to AAAI Press book Knowledge Discovery in Some exploration styles are formerly well enough
the Databases were excited at implicit benefits of this developed to have been made part of commercially
exploration. The editors hope that some of the excitement available software. Several expert system shells use
will communicate itself to the AI Magazine compendiums variations of ID3 for converting rules from
of this composition. exemplifications. Other systems use inductive, neural net, or
inheritable literacy approaches to discover patterns in
It has been estimated that quantum of information particular computer databases. Numerous forward- looking
in the world doubles every twenty months. The size and companies are using these and other tools to dissect their
databases count presumably increased indeed briskly. In databases for intriguing and useful patterns.
1989, total number of databases in the world is estimated at
5 million, although utmost of them are tiny DBASE III American Airlines searches its frequent leaflet
databases. The robotization of business conditioning database to find its better guests, targeting them for specific
produced an ever adding sluice of the data because indeed marketing elevations. Farm Journal analyzes its subscriber
simple deals, similar as a telephone call, use of the credit database and uses advanced printing technology to custom-
card, or the medical test, are generally recorded in figure hundreds of editions acclimatized to particular
computers. groups. Several banks, using patterns discovered in loan and
credit histories, have deduced better loan blessing and ruin
Scientific as well as government databases are also vaticination styles. General Motors is using a database of
fleetly growing. The National Aeronautics and Space machine trouble reports to decide individual expert systems
Administration had much further data than it dissect. Earth for colorful models. Packaged- goods manufacturers are
observation satellites, planned for 1990s, are anticipated to searching the supermarket scanner data to measure the
induce the one terabyte (1015 bytes) of data daily — further goods of their elevations and to look for shopping patterns.
than all previous operations combined. At a rate of 1 picture
each alternate, it would take the person several times A combination of business and exploration
(working nights/weekends) just to look at the film land interests has produced adding demands for, as well as
generated in each day. In biology, federally funded Human increased exertion to give, tools and ways for discovery in
Genome design will store the thousands of bytes for each of databases. This book is the first to bring together leading-
several billion inheritable bases. edge exploration from around the world on this content. It
spans numerous different approaches to discovery,
Closer to everyday lives, the 1990 U.S. tale data of including inductive literacy, Bayesian statistics, semantic
a million million bytes render the patterns that in retired query optimization, knowledge accession for expert
ways describe the cultures and mores of moment’s United systems, information proposition, and fuzzy sets.
States. What are we supposed to do with this deluge of
theraw data? Easily, little of it'll ever be seen by the mortal The book is aimed at those interested or involved
eyes. in computer wisdom and the operation of data, to both
inform and inspire farther exploration and operations. It'll be
Still, it'll have to be anatomized by the computers, of particular interest to professionals working with
if it can be understood at all. Although simple statistical databases and operation information systems and to those
ways for the data analysis were developed long ago, applying machine literacy to real- world problems.
advanced ways for the intelligent data analysis are not yet
mature. What Is Knowledge Discovery? There's an
immense diversity of current exploration on knowledge
As a result, there's a growing gap between the data discovery in databases. To give a point of reference for this
generation and data understanding. At the same time, there's
exploration, we begin then by defining and explaining Big data analytics may yield more important
applicable terms. analysis of issues ranging from mortality to case- reported
issues to resource application, and therefore could be more
Description of Knowledge Discovery clinically practicable. Machine literacy, for illustration,
evaluates patterns associated with an outgrowth directly
Knowledge discovery is the nontrivial birth of from the data, rather than from a pre-specified set of
implicit, preliminarily unknown, and potentially useful variables.
information from data. Given a set of data (data) F, a
language L, and some measure of certainty C, we define a A full range of associations and relations among
pattern as a statement S in L that describes connections the data are assessed. Whereas traditional statistical models
among a subset FS of F with a certainty c, similar that S is are done, machine literacy uses a training process whereby
simpler (in some sense) than the recitation of all data in FS. the model is iteratively given varied data sets to explore
numerous combinations of prophetic features to optimize
A pattern that's intriguing ( according to a stoner- analysis.
assessed interest measure) and certain enough ( again
according to the stoner’s criteria) is called knowledge. The Phenol-mapping, or deep phenol-typing, is another
affair of a program that monitors the set of data in a promising big data operation. Current complaint groups, or
database and produces patterns in this sense is discovered phenotypes, are squishy and miscellaneous Big data
knowledge. These delineations about the language, the analytics can identify analogous case clusters, creating
certainty, and the simplicity and interestingness measures multiple phenotypes within each complaint reality. In
are designedly vague to cover a wide variety of approaches. proposition, further refined phenol-mapping of complaint
Inclusively, these terms capture our view of the abecedarian countries and circles should help inform more customized-
characteristics of discovery in databases. In the following health opinions
paragraphs, we epitomize the connotations of these terms
and suggest their applicability to the problem of knowledge Big data can support the combination of multiple
discovery in databases. data sources from large case populations to more estimate
They noted that one of the foremost discovery the implicit benefits of curatives similar as ICD’s for
processes was encountered by Jonathan Swift’s Gulliver in individual cases. Indeed, big data logical styles are central
his visit to the Academy of Labado. The “ Design for to the success of perfection health, given the growing
perfecting academic Knowledge by practical and interest in incorporating data, which extensively increases
mechanical operations” was generating sequences of words the size and complexity of datasets. Similar datasets bear
by arbitrary permutations and “ where they plant three or advanced logical platforms and styles that are the emblems
four Words that might make Part of a Judgment, they of big data analytics.
mandated them to... Scribes.” This process, although
promising to produce numerous intriguing rulings in the Big data analysis can guide programs to address a
( veritably) long run, is rather hamstrung and was lately certain case member by specific interventions. The success
proved to be NP-hard. of the policy is critically dependent on the quality of the
underpinning exploration and the quality (effectiveness) of
the interventions. For numerous interventions (for case in
III. METHODOLOGY the social/ internal health sphere) widely accepted styles for
validating success are still lacking. There's several
A. BIG DATA ANALYTICAL MODEL challenges result regarding Big Data and population heart
complaint similar as
Big data analytics support the conception of
artificial intelligence at the heart of numerous new digital • Data protection regulation makes it delicate to dissect data
health platforms and perfection health tools. Immaculately, from different heart complaint providers and services in
application of big data logical tools in cardiovascular care combination
will restate into better care and issues at a lower cost.
• A significant part of the population health records is
The eventuality for more important prophetic unshaped heart complaint textbook
models is a charming operation of big data analytics.
Historically, prediction models have reckoned on a limited • There are interoperability, data quality and data integration
number of specified variables manually entered to estimate limitations
a‘ threat score’. Similar models generally warrant perfection .
they perform‘ nicely well at the population position, but not
at the individual patient position.
And despite the actuality of dozens of threat
models related to cardiovascular conditions, many are
employed to make remedial opinions.
 Data Set Collection
chest pain type 1.0 0.0

(Heat Dataset)
resting blood 130.0 140.0
UCI- Repository pressure
maximum heart 109.0 79.0
rate achieved
exercise induced 0.0 1.0
Convert Time Stream
Data angina
[Heart Beat Rate]

Table 3.1 Normalized Heart Dataset

Min/ Max Normalization Process C. GREEDY FEATURE EXTRACTION MODEL

Feature selection is one of the dimension reduction

ways which have been used to allow a better understanding
Feature Selection Algorithm of data and ameliorate the performance of other literacy
tasks. Although the selection of applicable features has been
considerably studied in supervised literacy, point selection
Classification Algorithm with the absence of class markers is still a grueling task.
(KNN, SVM, CNN)

This paper develops a new system for unsupervised

Metrics and Accuracy Analysis point selection, which efficiently selects features in a greedy
manner. The paper first defines an effective criterion for
unsupervised point selection which measures the
Predict Heart Disease Patient reconstruction error of the data matrix grounded on the
named subset of features. The paper also presents a new
algorithm for greedily minimizing the reconstruction error
grounded on the features named so far. The greedy
Fig 3.1 Architecture Diagram Of Proposed System algorithm is grounded on an effective recursive formula for
calculating the reconstruction error.
B. NORMALIZATION MODEL
The greedy algorithm selects replication the most
Data transformation such as Normalization is a representative point among the remaining features, and also
data preprocessing tool used in data mining system. An eliminates the effect of the named features from the data
attribute of a dataset is normalized by scaling its values so matrix. This step makes it less likely for the algorithm to
that they fall within a small-specified range, such as 0.0 to elect features that are analogous to preliminarily named
1.0. Normalization is particularly useful for classification features, which consequently reduces the redundancy
algorithms involving neural networks, or distance between the named features. In addition, the use of the
measurements such as nearest neighbor classification and recursive criterion makes the algorithm computationally
clustering. There are many methods for data normalization doable and memory effective compared to the state of the
includes min-max normalization, z-score normalization and art styles for unsupervised (forward and backward) point
normalization by decimal scaling. selection.
Min-max normalization performs a linear
transformation on the original heart dataset. Min-max D. CLASSIFICATION ANALYTICAL MODEL
normalization maps a value d of P to d′ in the range
[new_min(p), new_max(p)]. Machine Literacy indicates how computers can learn or
ameliorate their performance using data. System programs
Min max normalization preserves the relationship among to generally plant to descry the styles and prove specialized
predicated opinions on data. Machine knowledge is a fast
the original heart dataset values. The table 4.1 Describe a growing discipline. Then, using classic problems in machine
sample normalized heat disease dataset model details shows, literacy that are largely affiliated to data mining.
• Supervised bracket literacy
Attribute Original Normalized Supervised bracket literacy model correspond of all data is
Values dataset labeled and algorithm learn to prognosticate the affair from
training dataset.
age 70.0 35.0 E.g. SVM.
 • Unsupervised bracket learning Value 3:
Unsupervised bracket literacy is used for clustering downsloping )
grounded algorithm. In this session all the information is Major vessels (0-
mot included and algorithm founds to essential structure Ca 3) colored by Integer
flouroscopy
from the input dataset.
Result of thalium
E.g. K- means, KNN. Neural Networks stress test (Value
•Semi-supervised bracket Learning 3 = normal;
Semi-supervised literacy is a combination of supervised Thal Integer
Value 6 = fixed
literacy and unsupervised literacy. In Semi-supervised defect; Value 7 =
literacy some data is labeled and some data isn't labeled. In reversable defect )
this approach, labeled training dataset are used to learn class status of heart
models and unlabelled training dataset are used to define the disease
boundaries between classes. (angiographic
status)
Value 0: < 50%
IV. RESULTS AND DISCUSSIONS Num
diameter
Binary
A. DATASET DESCRIPTION narrowing
Value 1: > 50%
Preparing the database-for carrying the result, this paper diameter
used Heart patient data sets from ILPD (Indian Heart Case) narrowing
Data Set (table 4.1).
Table 4.1 Dataset Attribute
Table 4.1 describes the attribute type, description and
Attributes Gender Gender Categorical values.
Description
Type Categorical
age age given in years Real number
sex (Value 1 :
Sex male; Value 0 : String
female)
chest pain type(1:
typical angina ; 2:
atypical angina
Cp Real number
3: non-anginal
pain ;4:
asymptomatic)
resting blood
pressure (in mm
Trestbps Real number
Hg on admission
to the hospital)
Cholestoral(Seru
Chol m cholestoral) in Real number
mg/dl
Fasting blood
sugar in mg/dl
Fbs (>120) Real number Fig 4.1 Performance of SVM Classification
Value 1 = true;
Value 0 = false)
Resting
Restecg electrocardiograph Real number
ic results
Heart rate
Thalach achieved at Integer
maximum
Exercise induced
Exang angina (Value 1 : Integer
yes; Value 0 : no)
ST depression
originated by
Oldpeak Integer
exercise relative to
rest
Slope Slope of the peak Integer
exercise ST
segment (Value
1: upsloping ;
Value 2: flat ;
 [2] W.J. Frawley and G. Piatetsky-Shapiro, “Knowledge
Discovery in Databases: An Overview”, AI Magazine, Vol.
13, No. 3, pp. 57-70, 1996.
[3] Kiyong Noh, HeonGyu Lee, Ho-Sun Shon, Bum Ju Lee
and Keun Ho Ryu, “Associative Classification Approach for
Diagnosing Cardiovascular Disease”, Intelligent Computing
in Signal Processing and Pattern Recognition, Vol. 345, pp.
721-727, 2006.
[4] Latha Parthiban and R. Subramanian, “Intelligent Heart
Disease Prediction System using CANFIS and Genetic
Algorithm”, International Journal of Biological, Biomedical
and Medical Sciences, Vol. 3, No. 3, pp. 1-8, 2008.
[5] Sellappan Palaniappan and Rafiah Awang, “Intelligent
Fig 4.2 Performance of KNN Classification Heart Disease Prediction System using Data Mining
Fig 4.1, Fig 4.2 describes the performance analysis of SVM Techniques”, International Journal of Computer Science and
and KNN classification algorithms respectively.
Network Security, Vol. 8, No. 8, pp. 1-6, 2008
[6] Shantakumar B. Patil and Y.S. Kumaraswamy,

V. CONCLUSION “Intelligent and Effective Heart Attack Prediction System

using Data Mining and Artificial Neural Network”,
The project focuses on SVM classification
algorithms to effectively detect heart disease types. The European Journal of Scientific Research, Vol. 31, No. 4, pp.
dataset is taken from UCI repository. Preprocessed such as 642-656, 2009.
zero value, N/A value and unicode character elimination is
carried out here. Important features are extracted out for [7] Nidhi Singh and Divakar Singh, “Performance
better classification. Confusion matrix is prepared with Evaluation of K-Means and Hierarchal Clustering in Terms
accuracy score calculation. In addition, KNN classification
algorithms as well as neural network to effectively detect of Accuracy and Running Time”, Ph.D Dissertation,
risk types are also carried out. The dataset is taken and Department of Computer Science and Engineering,
preprocessed such as unicode removal. Important features
are extracted out for better classification. Confusion matrix Barkatullah University Institute of Technology, 2012.
is prepared with accuracy score calculation. Accuracy [8] Weiguo, F., Wallace, L., Rich, S., Zhongju, Z.: “Tapping
prediction is also carried out. Convolutional Neural
Network based prediction model is worked out to find the Power of Text Mining”, Communication of the ACM.
algorithm efficiency. 540 training records and 210 test 49(9), 77-82, 2006.
records are taken out for convolutional neural network
training. There are several directions for future research. [9] Jiawei Han M. K, Data Mining Concepts and
The current investigation of classification is still Techniques, Morgan Kaufmann Publishers, An Imprint of
preliminary.
Elsevier, 2006
REFERENCES [10] Huang Z, “Extensions to the k-means algorithm for
[1] Franck Le Duff, CristianMunteanb, Marc Cuggiaa and clustering large data sets with categorical values,” Data
Philippe Mabob, “Predicting Survival Causes After Out of Mining and Knowledge Discovery, Vol.2, pp:283–304,
Hospital Cardiac Arrest using Data Mining Method”, 1998
Studies in Health Technology and Informatics, Vol. 107,
No. 2, pp. 1256-1259, 2004. [11] A comparison of antiarrhythmic-drug therapy with
implant-able defibrillators in patients resuscitated from
near-fatalventricular arrhythmias. The Antiarrhythmics
versusImplantable Defibrillators (AVID) Investigators. N
Engl JMed. 1997 Nov 27;337(22):1576-83.
[12] R.Wu, W.Peters, M.W.Morgan, “The Next Generation
Clinical Decision Support: Linking Evidence to Best
Practice”, Journal of Healthcare Information Management.
16(4), pp. 50-55, 2002.
[13] Mary K.Obenshain, “Application of Data Mining
Techniques to Healthcare Data”, Infection Control and
Hospital Epidemiology, vol. 25, no.8, pp. 690–695, Aug.
2004.
[14] G.Camps-Valls, L.Gomez-Chova, J.Calpe-Maravilla,
J.D.MartinGuerrero, E.Soria-Olivas, L.Alonso-Chorda,
J.Moreno, “Robust support vector method for hyperspectral
data classification and knowledge discovery.” Trans.
Geosci. Rem. Sens. vol.42, no.7, pp.1530–1542, July.2004.
[15] Obenshain, M.K: “Application of Data Mining
Techniques to Healthcare Data”, Infection Control and
Hospital Epidemiology, 25(8), 690–695, 2004.
[16] Wu, R., Peters, W., Morgan, M.W.: “The Next
Generation Clinical Decision Support: Linking Evidence to
Best Practice”, Journal Healthcare Information
Management. 16(4), 50-55, 2002.
[17] Charly, K.: “Data Mining for the Enterprise”, 31st
Annual Hawaii Int. Conf. on System Sciences, IEEE
Computer, 7, 295-304, 1998.
[18] Blake, C.L., Mertz, C.J.: “UCI Machine Learning
Databases”, http://mlearn.ics.uci.edu/databases/heart-
disease/, 2004.
[19] Mohd, H., Mohamed, S. H. S.: “Acceptance Model of
Electronic Medical Record”, Journal of Advancing
Information and Management Studies. 2(1), 75-92, 2005.