CSEIT1723320 | Received : 12 June 2017 | Accepted : 21 June 2017 | May-June-2017 [(2)3: 872-875]
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2017 IJSRCSEIT | Volume 2 | Issue 3 | ISSN : 2456-3307
872
A Classification Approach for instant Medical Assistance to
Health Seekers
Pritha Tikariha
*
, Prashant Richhariya
Department of Computer Science, CSIT, Durg, Chhatishgarh, India
ABSTRACT
Data mining approach is applied in numerous numbers of fields for predicting and forecasting the events. However,
in healthcare sectors, due to lack of faith in prediction method people hesitate to utilize data mining technique for
health issues. People post their health related queries and get reply from the experts in many online healthcare
applications. However, health seekers do not get instant assistance there; they need to wait for the experts for their
opinion. Many data is accumulated in repository of such application. Using data mining techniques, useful
information can be extracted from such repository, which can help health seekers to get instant assistance for their
health related issues. These paper presents analysis on some data mining technique particularly in disease dataset.
Three classification algorithms i.e. KNN, SVM, Naïve Bayes are applied on three disease dataset, to analyse the
performance of the classifiers. The predictive rate is evaluated using four evaluation parameters i.e. Accuracy,
precision, recall and f_measure. The experiment is performed in Matlab tool shows that Naive Bayes outperforms as
compared to rest of the classifiers.
Keywords : Arduino, Wi-Fi (ESP 8266), Load cell, Database System
I. INTRODUCTION
In health care domain, data can be very valuable. These
data can be mined and converted into useful
information. Medical data mining provides a way to
explore the hidden relationship present in the data set
of the medical realm. This relationship can be used for
the diagnosis of many diseases. However, these
medical data sets are very huge and obscure. These
dataset have to be structure and assimilated to form a
medical information system. Medical data mining
provides a way to achieve these things.
Patients generally have different medical attribute as a
result they have heterogeneous medical requirement.
These heterogeneous attribute needs to be classified
accordingly and transformed into homogeneous groups.
For converting it into homogeneous groups, these
dataset require detailed, effective and efficient
classification algorithms. Homogeneity brings the
benefits of increased certainty in clinical diagnosis,
predicting individual patient needs and resource
utilization.
In these research work three classification algorithms is
applied on different disease dataset, to determine the
prediction rate of the classifier. KNN, SVM and Naïve
Bayes algorithm are analysed using for evaluation
parameters accuracy, precision, recall and f-measure.
The remaining paper is organized in the following way:
Section II reviews the previous work done on different
dataset and classifiers. Section III describes the
methodology and the dataset. Section IV includes the
experimental setup results and finally section V has
conclusion of the research work and the future scope.
II. LITERATURE SURVEY
Classification algorithms are generally very useful for
medicinal issues, especially when applied for the
diagnosis purpose [4]. Many machine learning
algorithms are applied in the medical domain in the
course of recent decades [5] [6] [7] [8] [9] [10].A large
portion of these applications are specific and include
machine learning procedure like using data mining for
diagnosis purpose [11], applied neural network rule for
the prediction of breast cancer [10]. Data mining has