International Journal of Reliable and Quality E-Healthcare

Volume 12 • Issue 2

DDPIS: Diabetes Disease Prediction

by Improvising SVM
Shivani Sharma, ABES Institute of Technology, India*
https://orcid.org/0000-0003-3381-269X

Bipin Kumar Rai, ABES Institute of Technology, India

https://orcid.org/0000-0002-9834-8093

Mahak Gupta, ABES Institute of Technology, India

Muskan Dinkar, ABES Institute of Technology, India

ABSTRACT

An illness that lasts longer and has continual repercussions is known as a chronic illness. Adults
all across the world die as a result of chronic sickness. Diabetes disease prediction by improvising
support vector machine is a platform that predicts diabetes based on the data entered into the system
and offers reliable results based on that data. Earlier, the dataset consisted of a smaller number of
features comprising the patients’ medical details that were useful in determining the patient’s health
condition and was mainly focused on gestational diabetes, which only deals with pregnant women.
In this work, the authors build a system that is more efficient than the previous system because of
these reasons. It provides more accurate results by improvising the support vector machine, which
includes more datasets and can predict the possibility of diabetes disease in both males and females.

Keywords
Accuracy, Diabetes Disease Prediction, Machine Learning, Support Vector Machine

INTRODUCTION

Diabetes is one of the most widespread and fatal chronic diseases that harm the entire body system.
The body of a diabetic patient has a high level of blood sugar (Lyngdoh et al., 2021). A person with
a chronic illness has a condition that lasts longer and has ongoing consequences. One of the most
significant disadvantages of chronic disorders is that they have a detrimental impact on people’s
standard of living. It is one of the most dangerous infections that may be discovered worldwide.
This chronic illness costs the lives of adults all over the world.(Ahmed et al., 2021; Lai et al., 2019).
Chronic diseases have a monetary burden attached to them and cost a lot of money for governments
and people. As we all know, the operation cost is high and not every family can afford it. Two factors

DOI: 10.4018/IJRQEH.318090 *Corresponding Author

This article published as an Open Access article distributed under the terms of the Creative Commons Attribution License
(http://creativecommons.org/licenses/by/4.0/) which permits unrestricted use, distribution, and production in any medium,
provided the author of the original work and original publication source are properly credited.

1
International Journal of Reliable and Quality E-Healthcare
Volume 12 • Issue 2

can cause diabetes: (1) the pancreas produces insufficient insulin, or (2) the body produces insufficient
insulin. Only 5–10% of people with diabetes have this type of disease (Type-1) or (2). The produced
insulin does not affect the cells (Type-2). Insulin is the hormone that controls the uptake of glucose
from the bloodstream into most cells (muscles and fat cells). If there isn’t enough insulin, glucose
won’t have the same effect as it usually does, and glucose won’t be absorbed by the body cells that
need it (Deberneh & Kim, 2021).
Diabetes mellitus is one of the leading causes of death in the United States. It requires
detection and diagnosis at an early stage. Diagnosis of diabetes and interpretation of diabetes
data is a significant categorization issue (Deberneh & Kim, 2021; Saeedi et al., 2019). Diabetes
also afflicted approximately 463 million people aged 20 to 79 in 2019. (International Diabetes
Federation-IDF) (Gulshan et al., 2016). Seventy-nine percent of the adult population live in low-
and middle-income countries. According to estimates (IDF), approximately 700 million people
will have diabetes by 2045 (Soni & Varma, n.d.). Every year, the number of instances grows,
and the number of active cases continues to rise. Diabetes has become one of the most severe
and rapid diseases to claim many people’s lives worldwide, so it is essential to be concerned
(Nayak & Pandi, 2021; Perveen et al., 2016). According to research, 70% of people in India suffer
from this widespread disease, and 25% die due to early ignorance. The primary motivation for
developing this project is so that a user can sit at their convenience and check their health (Vizhi
& Dash, 2020; Zhou et al., 2020).
We developed the platform diabetes disease prediction by improvising a support vector machine
to overcome diabetes disease in earlier stages. As we all know, in the competitive economic
development environment, people are so busy making money and improving their lifestyle and
future that they are not concerned about their health. The leading causes of ignorance are that they
do not have time. They are so busy with their work that they neglect their health and do not go for
regular body check-ups, which are essential for monitoring an individual’s health to be free from
any disease harmful to their body that may cost their life. People have become so preoccupied with
their daily lives that they have no time to schedule appointments and consult a doctor, resulting
in fatal conditions. Our diabetes prediction system helps individuals to predict the possibility of
diabetes without taking more of their time. Whenever they are free from work, they can immediately
check the likelihood of diabetes. They can consult the doctor for further treatment or assistance
if the results are positive.
Machine learning is a kind of artificial intelligence (AI) that lets software applications become
more accurate and efficient in predicting outcomes. ML algorithm uses historical data to anticipate
improved output values (Kaur, 2019; Kumar et al., 2022).
Support Vector Machine, i.e., SVM, is a machine learning algorithm based on supervised
learning. SVM can be used for classification and regression complications but mainly for classification
problems. The main aim of the support vector machine is to find a hyperplane in n-dimensional space
(where n is the total number of attributes). The dimension in the hyperplane depends on the number
of attributes used (Pranto et al., 2020; Rani, 2020)
Let’s consider an example where we have two independent variables x1, x2 and one of them
is dependent on either the blue or red. From the first figure, we now have to choose the best line to
segregate our data points.(Shafi & Ansari, 2021)
We choose the hyperplane whose distance from it to the nearest data point on each side is
maximized. If such a hyperplane exists, it is known as the maximum-margin hyperplane/hard margin.
So, from the above figure, we choose L2.

LITERATURE REVIEW

Arwatki Chen Lyngdoh et al. (2021) compared machine learning algorithms such as KNN, SVM,
DT, RF, and Naive Bayes. They compared all the classifiers and obtained the highest accuracy of

2
 International Journal of Reliable and Quality E-Healthcare
Volume 12 • Issue 2

Figure 1. SVM before finding hyperplane and after finding hyperplane. (Lyngdoh et al., 2021)

76% with KNN, and with the other remaining classifiers, they got above 70% accuracy(Lyngdoh
et al., 2021).
Hang Lai et al. (2019) used machine learning techniques such as LR and Gradient Boosting
Machine (GBM) to predict the occurrence of diabetes mellitus. They obtained the Area under the
Receiver Operating Characteristic Curve (AROC) for the GBM was 84.07%, and for the LR model,
it was 84%. They also compared these models with other techniques, such as DT (80.5%) and RF
(83.04%) and found that the GBM and LG were more efficient (Lai et al., 2019).
Henock M. Deberneh et al. (2021) proposed a system that can predict Type 2 diabetes. For this
study, they collected the dataset from the private medical institute as electronic health records from
2013 to 2018. They used SVM, XG Boost, RF, LR, and ensemble classifiers and got an accuracy of
73%, 72%, 73%, and 71%, respectively (Deberneh & Kim, 2021).
Mitushi Soni et al. (2019) used different machine learning classifications to predict diabetes
disease. They use SVM, LR, KNN, Gradient Boosting Classifiers, DT, and RF to improve the
performance, which helps them increase the prediction model’s accuracy. The technique which
provided the highest accuracy compared to other machine learning techniques was Random Forest
(RF), with 77% accuracy.
N. Sneha et al. (2019) focused on analyzing diabetes using optimal feature selection. They use
various algorithms such as SVM, RF, NB, DT and KNN and get an accuracy of 77.73, 75.39, 73.48,
73.18, and 63.04%, respectively (Sneha & Gangil, 2019). Lomani Nayak et al. (2021) applied three
algorithms to the Pima Indian Diabetes dataset, KNN, SVM, and Decision Tree, to predict early
diabetes. They also compared the SVM with the other two algorithms and got the highest accuracy
of 73.95%, whereas KNN provided 71.35% and DT provided 72%.
Sajida et al. (2016) evaluated data mining classification techniques and their performance for
analysis. The Canadian Primary Care Sentinel Surveillance Network (CCPCSSN) was the dataset used

3
International Journal of Reliable and Quality E-Healthcare
Volume 12 • Issue 2

in this paper. They used AdaBoost, a bagging ensemble technique using the J48 decision tree, to classify
diabetic patients across three age groups, which were 18–35, 36–55, and older than 55. They found that
the AdaBoost ensemble method is better than bagging and the J48 decision tree (Perveen et al., 2016)
Hauping Zhou et al. (2020) proposed a system that can predict diabetes by using DNN. This
system can also determine the type of diabetes. They achieved 94.02% accuracy with the diabetes
type dataset and 99.41% accuracy with the Pima Indian Diabetes Dataset.
Kayal Vizhi et al. (2020) used different machine learning techniques such as KNN, SVM, LR,
DT, GNB, RF, and XG Boost to predict diabetes disease. They used the PIMA dataset and achieved
the highest accuracy of 77.64% with the LR.
Talha Mahboob Alam et al. (2019) aimed to predict the early prediction of diabetes. They used
ANN, RF, and K-means clustering techniques. The dataset they used was taken from the National
Institute of Diabetes and Digestive and Kidney Diseases. The K-means clustering achieved an accuracy
of 73.6%, RF completed 74.7%, and the highest, 75.7%, was achieved by ANN.
Kannadasan et al. (2021) focused on deep neural network (DNN) classifiers to predict diabetes.
They categorized diabetes using the SoftMax layer to extract the excellent features. They used
stacked autoencoders. The dataset that they used was the PID dataset, and they achieved an accuracy
of 86.26%. (Butt et al., 2021)
Anuja Kumari et al. (?) classified diabetes from a high dimensional dataset using SVM and
obtained an accuracy of 75% (Jegan et al., n.d.).
Jobeda et al.(2021) compared the seven ML algorithms to predict diabetes, also built a NN
model, and found out two hidden layers in NN give the best accuracy of 88.6%. (Khanam & Foo,
2021) Bharath et al. used the PIMA dataset on convolutional long short-term memory (CLSTM) Deep
Learning Technique to detect the occurrence of diabetes disease (Bharath et al., n.d.).
Yazan Jian et al. (2021) used a dataset from the Rashid Center for Diabetes and Research, which
is situated in UAE and applied ML to predict diabetes disease. (Jian et al., 2021)
Finally, Xue et al. (2020) used SVM, NB and Light GBM, collected datasets from UCI ML
Repository, and achieved the best accuracy with SVM. (Xue et al., 2020)
In our work, we found that most diabetes disease predictions are based on gestational diabetes,
which is present in pregnant ladies. The database used to train this system was the PIMA dataset,
which only contains attributes regarding female patients. The accuracy of SVM obtained by the
PIMA dataset was less.

METHODOLOGY

Data Collection
The dataset was collected from the UCI ML repository, which contains 16 attributes of 520 patients.
This dataset includes information on both female and male patients. Further, we’ve converted string
values such as YES or NO to binary values 0 and 1, where 0 means NO or Negative and 1 means
YES or Positive. Also, the gender binary value is 0 for Males and Females is 1.

Data Pre-processing
We used data pre-processing to make the dataset serviceable and obtain an understanding. We analyzed
the dataset for uncommon entries and fixed them manually to deal with erroneous records (Sharma et
al., 2022). To make a helpful dataset, we’ve used Pandas and NumPy library to deal with the dataset
efficaciously (Bano et al., 2021). We’ve converted string values such as YES or NO to binary values
0 and 1 where 0 means NO or Negative and 1 means YES or Positive. Also, the gender binary value
is 0 for Males and Females is 1.

4
 International Journal of Reliable and Quality E-Healthcare
Volume 12 • Issue 2

Table 1. List of attributes

S No. Attributes
1 Gender
2 Visual Blurring
3 Polydipsia
4 Delayed Healing
5 Genital Thrush
6 Partial Paresis:
7 Muscle Stiffness
8 Alopecia
9 Irritability
10 Itching
11 Sudden Weight Loss
12 Obesity
13 Weakness
14 Polyphagia
15 Polyuria
16 Age

Setting Classification Metrics

To categorize the disease and get a final result, we need to set a few metrics to help us predict
diabetes. Since we used the Sk-learn machine learning library (Jakka & Vakula Rani, 2019) for our
experiment, we’ve used the confusion matrix as the classification measure metrics. In our analysis,
the used metric, i.e., accuracy is listed below (Sahoo et al., 2020).
Accuracy (A) is defined as follows.

Tp + Tn
A= (1)
Tp + Tn + Fp + Fn

Precision represents the number of true positives correctly identified as diabetic patients over
the total number of positive predictions.
Precision (P) is defined as follows.

Tp
P= (2)
Tp + Tn

Architecture Diagram
An architecture diagram is used to describe the dynamic aspects of the system. The activity can
be described as an operation of the system. In this diagram, the activity starts from the user,
where the user registers into the system, logs in using the credentials and then the credentials are
matched into the system. If true, the user proceeds to the input phase, where the user enters input,

5
International Journal of Reliable and Quality E-Healthcare
Volume 12 • Issue 2

and then moves to the prediction phase, where the input is analyzed. Finally, after processing the
data from the datasets, the analysis will happen, and the correct result will be displayed, which is
nothing but the Output.
The system will detect whether the user or the person has diabetes disease or not. It gives the Output
in the form of YES or NO. When using the SVM algorithm for predicting the disease, the user enters
his credentials and answers some questions in yes or no terms. After that, the values are processed using
the SVM algorithm, and after this process, the Output is predicted in terms of Yes or NO.

Dataset Collection: The data is collected from the UCI dataset. The dataset has 16 attributes of
520 patients.
Data Pre-processing: This is the most critical process. It is used to improve the efficiency and quality
of data. Data pre-processing is done in two steps which are as follows:
Missing Value Removal: As our dataset does not contain any missing values, the process for removing
missing values will be skipped.
Splitting The Data: When data cleaning is done, the data will get normalized in the training and
testing model.

After splitting the data, it is trained using logic and an algorithm.

Applying the SVM Algorithm: We use the SVM algorithm after the data is pre-processed. The
algorithm is applied to UCI datasets and analyzes the algorithm’s accuracy.

On the User end, the following activity will take place:

Login and Registration: The user or patient will log into the system, enter their details, and
register themselves.
Enter The Details: The user will enter the details such as age and gender and the symptoms in terms
of yes or no.
Match Values: The values the user enters are now matched with the database by applying the
SVM algorithm.
Output Is Generated: After matching the values and applying the SVM algorithm, the Output
is generated, and the result is displayed to the user. The Output is presented in terms of
YES or NO.

Figure 2. Architecture diagram

6
 International Journal of Reliable and Quality E-Healthcare
Volume 12 • Issue 2

Workflow
A Workflow is a type of diagram representing a system’s process. It can be defined as a diagrammatic
representation of a diabetes disease prediction system, a step-by-step approach to predicting the
possibility of diabetes.
Initially, the data collection of diabetic patients is done. After that, the collected data moves to
the pre-processing stage, and irreverent features are removed from the datasets. After that, it turns to
the testing and training phase. When data testing and training are done, SVM algorithms are applied
to the data, and the predicted outcome is generated using SVM algorithms.
The workflow diagram defines various steps such as data collection, pre-processing, testing,
analysis and prediction. The data of diabetic patients is collected here from the UCI (University of
California Irvine) Machine Learning Repository, which is available for males and females. The UCI
dataset has 16 attributes, which will further help improve the system’s accuracy. The next step is
Data Pre-processing, where the raw data is manipulated and converted into efficient and valuable
data, increasing the system’s performance. That is preparing the raw data and making it suitable for
a machine-learning model. When data pre-processing is complete, we move to the next step, training
and testing the data. In the training model, the UCI Dataset is fed to the SVM algorithm to train the
model. It helps the program to understand the dataset for predicting the Output. Training datasets are
provided to machine learning algorithms to teach them how to make predictions or perform a desired
task. Now, the test data are the data which will determine whether our system returns the expected
result or not. Data testing measures performance, such as the algorithm’s accuracy. As the training
and testing of the data are done, we apply the SVM algorithm to the available data, which will help
predict diabetes. The SVM algorithm is used to help predict the possibility of diabetes and provide
the user with an output.

IMPLEMENTATION AND RESULTS

1. Import the dependencies

2. Load the diabetes dataset to a pandas DataFrame.
3. Standardize the data using the function scaler.transform()
4. Split the dataset by using train_test_split()
5. Use function SVC(kernel = ‘linear’)
6. Use Classifier. fit() function to train the model.
7. Perform prediction on the test set using Classifier. predict()

We’ve made a model to predict diabetes with 520 classes and 16 attributes; among them, 320 are
marked as 1, i.e., Positive and 200 are marked as 0, i.e., Negative.
Heatmap can be defined as the graphical representation of data using various colors to create the
value of the matrix. The darker colors represent the higher values, whereas the brighter one represents
the low value of multiple attributes in the below figure.

Figure 3. Workflow diagram

7
International Journal of Reliable and Quality E-Healthcare
Volume 12 • Issue 2

Figure 4. Outcome count

Figure 5. Heat map

8
 International Journal of Reliable and Quality E-Healthcare
Volume 12 • Issue 2

Accuracy Score
The accuracy we got from our Diabetes Disease Prediction by Improvising SVM is 93.26%. We have
used 16 attributes to improvise the performance of SVM, including the attributes of both females
and males.

CONCLUSION

Diabetes is a fatal chronic disease that harms the entire body system. The body of a diabetic patient
has a high level of blood sugar. Various machine learning techniques could be utilized to forecast the
presence of disease, such as SVM, Logistic regression, KNN, XGBoost, etc. In our research paper, we
propose a diabetes occurrence prediction system that can predict the occurrence of diabetes disease
using SVM. Earlier, the dataset consisted of a smaller number of features comprising the patients’
medical details that were useful in determining the patient’s health condition. It was mainly focused
on gestational diabetes. In this implementation, we used a dataset comprising more features, which
helped us increase the accuracy of SVM to 93.26%. The dataset included females and males, and it
was built to help patients assess the risk of diabetes.

ACKNOWLEDGMENT

Competing Interests
All authors of this article declare there are no competing interest.

Funding Agency
This research received no specific grant from any funding agency in the public, commercial, or not-
for-profit sectors. Funding for this research was covered by the authors of the article.

9
International Journal of Reliable and Quality E-Healthcare
Volume 12 • Issue 2

REFERENCES

Ahmed, N., Ahammed, R., Islam, M., Uddin, M. A., Akhter, A., Talukder, M. A., & Paul, B. K. (2021). Machine
learning based diabetes prediction and development of smart web application. International Journal of Cognitive
Computing in Engineering, 2, 229–241. doi:10.1016/j.ijcce.2021.12.001
Bano, F., & Munidhanalakshmi, K. (2021). Predict Diabetes Mellitus Using Machine Learning Algorithms. Journal
of Physics: Conference Series, 2089(1), 012002. Advance online publication. doi:10.1088/1742-6596/2089/1/012002
Bharath, P., Chowdary, K., & Udaya Kumar, R. (n.d.). An Effective Approach for Detecting Diabetes using Deep
Learning Techniques based on Convolutional LSTM Networks. International Journal of Advanced Computer
Science and Applications, 12(4). www.ijacsa.thesai.org
Butt, U. M., Letchmunan, S., Ali, M., Hassan, F. H., Baqir, A., & Sherazi, H. H. R. (2021). Machine Learning
Based Diabetes Classification and Prediction for Healthcare Applications. Journal of Healthcare Engineering,
2021, 1–17. Advance online publication. doi:10.1155/2021/9930985 PMID:34631003
Deberneh, H. M., & Kim, I. (2021). Prediction of type 2 diabetes based on machine learning algorithm.
International Journal of Environmental Research and Public Health, 18(6), 3317. Advance online publication.
doi:10.3390/ijerph18063317 PMID:33806973
Gulshan, V., Peng, L., Coram, M., Stumpe, M. C., Wu, D., Narayanaswamy, A., Venugopalan, S., Widner, K.,
Madams, T., Cuadros, J., Kim, R., Raman, R., Nelson, P. C., Mega, J. L., & Webster, D. R. (2016). Development
and validation of a deep learning algorithm for detection of diabetic retinopathy in retinal fundus photographs.
Journal of the American Medical Association, 316(22), 2402–2410. doi:10.1001/jama.2016.17216 PMID:27898976
Jakka, A., & Vakula Rani, J. (2019). Performance evaluation of machine learning models for diabetes prediction.
International Journal of Innovative Technology and Exploring Engineering, 8(11), 1976–1980. doi:10.35940/
ijitee.K2155.0981119
Jegan, C., Kumari, V. A., & Chitra, R. (n.d.). Classification Of Diabetes Disease Using Support Vector Machine
Identification and Rectification of Security Issues in IOT View project Development of an Intelligent system for
the diagnosis of cardiovascular diseases View project Classification Of Diabetes Disease Using Support Vector
Machine. https://www.researchgate.net/publication/320395340
Jian, Y., Pasquier, M., Sagahyroon, A., & Aloul, F. (2021). A machine learning approach to predicting
diabetes complications. Healthcare (Switzerland), 9(12), 1712. Advance online publication. doi:10.3390/
healthcare9121712 PMID:34946438
Kaur, H. (2019). Prediction of Diabetes Using Support Vector Machine. International Journal for Research in
Engineering Application & Management, 5, 2454–9150. doi:10.35291/2454-9150.2019.0076
Khanam, J. J., & Foo, S. Y. (2021). A comparison of machine learning algorithms for diabetes prediction. ICT
Express, 7(4), 432–439. doi:10.1016/j.icte.2021.02.004
Kumar, A., Goyal, A., Rai, B. K., & Sharma, S. (2022). OCR based medical prescription and report analyzer.
Proceedings of the International Conference on Computational Intelligence and Computing Applications-21
(ICCICA-21), 2424. 10.1063/5.0081176
Lai, H., Huang, H., Keshavjee, K., Guergachi, A., & Gao, X. (2019). Predictive models for diabetes mellitus using
machine learning techniques. BMC Endocrine Disorders, 19(1), 101. Advance online publication. doi:10.1186/
s12902-019-0436-6 PMID:31615566
Lyngdoh, A. C., Choudhury, N. A., & Moulik, S. (2021). Diabetes Disease Prediction Using Machine Learning
Algorithms. Proceedings - 2020 IEEE EMBS Conference on Biomedical Engineering and Sciences, IECBES
2020, 517–521. doi:10.1109/IECBES48179.2021.9398759
Mahboob Alam, T., Iqbal, M. A., Ali, Y., Wahab, A., Ijaz, S., Imtiaz Baig, T., Hussain, A., Malik, M. A., Raza,
M. M., Ibrar, S., & Abbas, Z. (2019). A model for early prediction of diabetes. Informatics in Medicine Unlocked,
16, 100204. Advance online publication. doi:10.1016/j.imu.2019.100204
Nayak, L., & Pandi, G. S. (2021). Diabetes Disease Prediction using Machine Learning. International Research
Journal of Engineering and Technology. www.irjet.net

10
 International Journal of Reliable and Quality E-Healthcare
Volume 12 • Issue 2

Perveen, S., Shahbaz, M., Guergachi, A., & Keshavjee, K. (2016). Performance Analysis of Data Mining Classification
Techniques to Predict Diabetes. Procedia Computer Science, 82, 115–121. doi:10.1016/j.procs.2016.04.016
Pranto, B., Mehnaz, S. M., Mahid, E. B., Sadman, I. M., Rahman, A., & Momen, S. (2020). Evaluating machine
learning methods for predicting diabetes among female patients in Bangladesh. Information (Switzerland), 11(8),
374. Advance online publication. doi:10.3390/info11080374
Rani, K. J. (2020). Diabetes Prediction Using Machine Learning. International Journal of Scientific Research
in Computer Science, Engineering and Information Technology, 294–305. 10.32628/CSEIT206463
Saeedi, P., Petersohn, I., Salpea, P., Malanda, B., Karuranga, S., Unwin, N., Colagiuri, S., Guariguata, L., Motala,
A. A., Ogurtsova, K., Shaw, J. E., Bright, D., & Williams, R. (2019). Global and regional diabetes prevalence
estimates for 2019 and projections for 2030 and 2045: Results from the International Diabetes Federation
Diabetes Atlas, 9th edition. Diabetes Research and Clinical Practice, 157. doi:10.1016/j.diabres.2019.107843
Sahoo, J., Dash, M., & Pati, A. (2020). Diabetes Prediction Using Machine Learning Classification Algorithms.
International Research Journal of Engineering and Technology. www.irjet.net
Shafi, S., & Ansari, G. A. (2021). Early Prediction of Diabetes Disease & Classification of Algorithms
Using Machine Learning Approach. SSRN Electronic Journal. 10.2139/ssrn.3852590
Sharma, S., Kesarwani, A., Maheshwari, S., & Rai, B. K. (2022). Federated Learning for Data Mining in
Healthcare. EAI/Springer Innovations in Communication and Computing. doi:10.1007/978-3-030-85559-8_16
Shrestha, R., & Chatterjee, J. M. (n.d.). Heart Disease Prediction System Using Machine Learning. LBEF
Research Journal of Science, 115.
Sneha, N., & Gangil, T. (2019). Analysis of diabetes mellitus for early prediction using optimal features selection.
Journal of Big Data, 6(1), 13. Advance online publication. doi:10.1186/s40537-019-0175-6
Soni, M., & Varma, S. (n.d.). Diabetes Prediction using Machine Learning Techniques. www.ijert.org
Vizhi, K., & Dash, A. (2020). Diabetes Prediction Using Machine Learning. International Journal of Advanced
Science and Technology, 29(6), 2842–2852.
Xue, J., Min, F., & Ma, F. (2020). Research on diabetes prediction method based on machine learning. Journal of
Physics: Conference Series, 1684(1), 012062. Advance online publication. doi:10.1088/1742-6596/1684/1/012062
Zhou, H., Myrzashova, R., & Zheng, R. (2020). Diabetes prediction model based on an enhanced deep neural
network. Eurasip Journal on Wireless Communications and Networking, 2020(1). 10.1186/s13638-020-01765-7

Shivani Sharma is working as Assistant Professor (IT) in ABESIT Ghaziabad. She is pursuing her PhD and have
completed M.Tech & B.Tech in Information Technology. She has more than 15 years of experience in different
renowned institution. Her area of interest are Machine Learning, Deep Learning & Software Testing. She has
published 10 research papers in different conferences (Scopus indexed), two book chapters in Springer. She has
also worked as reviewer of several Scopus indexed journal.

Bipin Kumar Rai, Ph.D. from Banasthali University, Rajasthan and M.Tech. & B.Tech. in Computer Science and
Engineering, is working as Professor (IT) in ABESIT Ghaziabad. Prof. (Dr.) Bipin Kumar Rai has more than 17
years of teaching experience in different renowned Institutions. His areas of interest are Cryptography & Information
Security, Blockchain, Compiler Construction, and Data Structures. He has published his Ph.D thesis work entitled
“Pseudonymization Based Mechanism for Security & Privacy of Healthcare: PcPbEHR Solution for Healthcare”
and M. Tech. dissertation work entitled “An Optimized Solution for Certified e-mail with Trusted Third Party”. He
has published 10 research papers in ESCI/Scopus indexed journals, 11 research papers in different Conferences
(Scopus indexed), 6 book chapters in Springer/CRC Press Taylor & Francis Group. He has worked as a Guest
Editor/Reviewer of several SCI/Scopus Indexed Journals.

Muskan Dinkar obtained her B.Tech (Information Technology) degree from ABES Institute of Technology, Ghaziabad
affiliated to Dr. A.P.J. Abdul Kalam Technical University and want to pursue career in Software Developing.

11