DIABETES PREDICTION USING DATA

SCIENCE IN PYTHON
Submitted in partial fulfillment of the requirements for the award
of

Bachelor of Computer Science

By
Name: NAVEENRANGASAMY B
(REGISTER NO 40290060)

JUSTCE PEACEREVOLUTTION

DEPARTMENT OF COMPUTER SCIENCE AND ENGINEERING

SCHOOL OF COMPUTING

SATHYABAMA
INSTITUTE OF SCIENCE AND TECHNOLOGY
(DEEMED TO BE UNIVERSITY)
Accredited with Grade "A" by NAAC
JEPPIAAR NAGAR, RAJIV GANDHI SALAI, CHENNAI 600199
MAY 2023
 SATHYABAMA
INSTITUTE OF SCIENCE AND TECHNOLOGY
(DEEMED TOBE UNIVERSITY)
Accredited "A"Grade by NAAC| 12B Status by UGC |Approved by AlCTE
www.sathyabama.ac.in
DEPARTMENT OF COMPUTER SCIENCE
BONAFIDE CERTIFICATE
This is to certify that this Project Report is the bonafide work of
NAVEENRANGASAMY B (40290060) who carried out the project entitled
"DIABETES PREDICTION USING DATA SCIENCE IN PYTHON" under my
supervision from January 2023 to May 2023.

INTERNAL GUIDE
Dr. V.ULAGAMUTHALVI M.E., Ph.D.
en
Dean and Head of the Department
Dr. REKHA CHAKRAVARTHIM.E., Ph.D.

Submitted for Viva Voce examination held on oE|202 3

, 23

Internal Examiner External Examiner

 DECLARATION

INAVEENRANGASAMY B hereby decare that the Project Report entitled

"DIABETES PREDICTION USING DATA SCIENCE IN PYTHON" done by me
under the guidance of Dr.V.ULAGAMUTHALVI M.E., Ph.D. at Sathyabama
Institute of Science and Technology is submitted in partial fulfillment of the
requirements for the award of Bachelor of Computer Science.

DATE: 08.05.20D23
PLACE: CHENNAI SIGNATURE OF THE CANDIDATE
 ABSTRACT

This is a classification problem of supervised machine learning. The objective is to

predict whether or not a patient has diabetes, based on certain diagnostic
measurements included in the dataset. Diabetes is a common chronic disease and
poses a great threat to human health. The characteristic of diabetes is that the
blood glucose is higher than the normal level , which is caused by defective insulin
secretion. Diabetes can lead to chronic damage and dysfunction of various tissues,
especially eyes, kidneys, heart, blood vessels and nerves. With the development
in living standards, diabetes is increasingly common in our day-to-day life.

Therefore, how to quickly and accurately diagnose and analyze diabetes is a topic
worthy studying. Machine learning can help people make a preliminary judgment
about diabetes according to their daily physical examination data, and it can serve
as a reference for doctors . For machine learning the selection of valid features and
the correct classifier are the most important problems. So in this study,
LogisticRegression and DecisionTree are implemented to predict the diabetes.

V
 TABLE OF CONTENTS

Chapter No. Title. Page No.

Abstract V

List of Figures VIII

1. INTRODUCTION

1.1 INTRODUCTION TO MACHINE LEARNING 10

USING PYTHON

1.2 NEED FOR THE MACHINE LEARNING 11

1.3 MACHINE LEARNING ALGORITHMS 12

1.4 INTRODUCTION TO PYTHON FOR DATA
14
ANALYTICS
14
1.5 ANACONDA
1.6 ANACONDA NAVIGATOR 15
1.7 JUPYTER NOTEBOOK
16
2. LITRATURE SURVEY 19

3. AIM AND SCOPE OF THE PRESENT

INVESTIGATION
3.1 AIM 25
3.2 SCOPE
3.3 WHAT IS DIABETES

4. EXPERIMENTS OR REQUIRMENTS AND 27

METHODS USED

VI
 4.1 SYSTEM SPECIFICATION 27
4.2 PROJECT DESCRIPTION 27
4.3 LIBRARIES USED 27
4.4 LOGISTIC REGRESSION 29

5. RESULTS AND DISCUSION, PERFORMANCE 40

ANALYSIS

6. SUMMARY AND CONCLUSION 42

APPENDIX
43
A. SOURCE CODE
46
B. SCREEN SHOTS

REFERENCES 52

VII
 LIST OF FIGURES

FIGURE DESCRIPTION PAGE NO.

NO.

4.1 System Architecture 29

4.2 Project Implementation 30

4.3 Code of importing libraries 31

4.4 Code for importing data set 32

4.5 Example code for the prep-work of data set with 33

null values

4.6 Example code for the prep-work of data set with 33

null values

4.7 Code for splitting data frame into two parts based 34
on the column values

4.8 Code for splitting data frame into two parts based 35
on the column values

4.9 Splitting Datasets 35

4.10 Code to demonstrate fitting the algorithm to the 36

model and resulting array of predictions made

4.11 Code to calculate the accuracy of the model 37

4.12 Accuracy graph model 37

4.13 Code for importing metrics to find the accuracy 38

score

VIII
5.1 Code for the prediction model 40

6.1 Code of importing libraries 46

6.2 Code for importing data set 46

6.3 Example code for the prep-work of data set with 47

null values

6.4 Example code for the prep-work of data set with 47

null values

6.5 Code for splitting data frame into two parts based 48
on the column values

6.6 Code for splitting data frame into two parts based 48
on the column values

6.7 Code to demonstrate fitting the algorithm to the 49

model and resulting array of predictions made

6.8 Code to calculate the accuracy of the model 50

6.9 Accuracy graph model 50

6.10 Code for importing metrics to find the accuracy 51

score

6.11 Code for the prediction model 51

IX
 CHAPTER 1
INTRODUCTION

This is a classification problem of supervised machine learning. The objective is to

predict whether or not a patient has diabetes, based on certain diagnostic
measurements included in the dataset. Diabetes is a common chronic disease and
poses a great threat to human health. The characteristic of diabetes is that the
blood glucose is higher than the normal level, which is caused by defective insulin
secretion. Diabetes can lead to chronic damage and dysfunction of various tissues,
especially eyes, kidneys, heart, blood vessels and nervesWith the development in
living standards, diabetes is increasingly common in our day-to-day life. Therefore,
how to quickly and accurately diagnose and analyze diabetes is a topic worthy
studying. Machine learning can help people make a preliminary judgment about
diabetes according to their daily physical examination data, and it can serve as a
reference for doctors . For machine learning the selection of valid features and the
correct classifier are the most important problems. So in this study,
LogisticRegression and DecisionTree are implemented to predict the diabetes.

1.1Introduction to MACHINE LEARNING USING PYTHON

Machine learning is a type of artificial intelligence (AI) that provides computers with
the ability to learn without being explicitly programmed. Machine learning focuses
on the development of Computer Programs that can change when exposed to new
data. In this article, we’ll see basics of Machine Learning, and implementation of a
simple machine-learning algorithm using python.
Machine learning is a method of teaching computers to learn from data, without
being explicitly programmed. Python is a popular programming language for
machine learning because it has a large number of powerful libraries and
frameworks that make it easy to implement machine learning algorithms.
Python offers multiple great graphing libraries that come packed with lots of
different features. No matter if you want to create interactive, live or highly

10
customized plots python has an excellent library for you.

To get a little overview here are a few popular plotting libraries:

MATPLOTLIB: It is a multi-platform data visualization library built on NumPy

arrays, and designed to work with the broader SciPy stack. It was conceived by
John Hunter in 2002, originally as a patch to I Python for enabling interactive
MATLAB-style plotting via gnu plot from the I Python command line.

PLOTLY: Plotlib is the Python Library for interactive data visualizations. Plt allows
you to plot superior interactive graphs than either Matplotlib or Seaborn.

1.2 Need for the MACHINE LEARNING

The field of artificial intelligence known as machine learning is concerned with the
process by which computers attempt to forecast future events using historical data
and the information they already possess. There are two distinct forms of machine
learning. The first kind of learning is called supervised learning, and in this type of
learning, the data itself serve as the instructor, and the system is constructed
based on the dataset. The second kind of learning is called unsupervised learning,
and it involves the data teaching itself by identifying certain patterns within the
dataset and then categorising those patterns. Over the last several years, a large
number of writers have reported and discussed their research on diabetes
prediction by utilising machine learning algorithms.

11
1.3 MACHINE LEARNING ALGORITHM

The research that has been done on machine learning has resulted in the
development of multiple data mining methods. These algorithms may be directly
applied to a dataset in order to create some predictions or to derive important
inferences and conclusions from such a dataset. Immediate use of these
algorithms is possible. Decision tree,NaivE Bayes, k-means, neural network, and
other similar algorithms are examples of prominent data mining techniques. In the
part that comes after this one, we will talk about them.

The Naive Bayes (NB) method is a straightforward approach to the construction

of classifiers. It is a Bayesian probabilistic classifier that uses Bayes' theorem as
its foundation. Given the class variable, each and every Naive Bayes classifier
operates on the assumption that the value of a single feature does not rely in any
way on the value of any additional feature. The following statement illustrates
Bayes theorem: Here X is the data tuple and C is the class, P(C|X) is calculated
by multiplying P(X|C) by P(C)/P(X). This ensures that P(X) remains the same for
all classes. Despite the fact that it operates on the assumption that feature values
are critically independent, which is not a realistic assumption, it performs very well
on huge datasets when this requirement is assumed and is true.

A decision tree is a kind of decision support system that utilises a tree-like

structure or representation of actions and their potential repercussions, which may
include the results of chance events as well as utility considerations. It is one of
the methods in which an algorithm may be shown. In the field of operational
research, decision trees are often used, particularly in the process of decision
analysis, to assist with the identification of a strategy that would most likely result
in the achievement of the objective. In the field of machine learning, it is also a
widely used tool. The process of mapping first from tree's parent node to each of
its leaf nodes individually may quickly and efficiently convert a decision tree to a
collection of rules. When everything is said and done, reasonable judgments may

12
be obtained by adhering to these criteria.

Support Vector Machine (SVM) - an abbreviation for It is a kind of learning called

supervised learning, and it divides data into two categories using a hyper plane as
the dividing line. The Support Vector Machine (SVM) accomplishes the same goal
as C4.5, with the exception that it does not make any use of Decision Trees. In
order to reduce the likelihood of an incorrect classification being made, the support
vector machine makes an effort to increase the margin, which is the distance
across the hyper plane and the 2 data points that are closest to it from each class.
Scikit-learn, MATLAB, and LIBSVM are examples of well-known software
packages that may be used to create support vector machines.

A Network Made of Artificial Neurons (ANN) A computer model that replicates

the architecture and functionality of biological neurons is referred to as an artificial
neural network, or ANN for short. Since a neural network adapts or trains in some
manner depending on the outputs and inputs, for that specific stage and
subsequently for each phase, data that travels thru the network has an effect on
the architecture of the ANN. ANNs are recognised as nonlinear statistical data
modelling software because of their ability to simulate the intricate interactions that
exist across inputs and outputs or to identify trends. ANNs are built using layers
that are linked to one another. The artificial neural networks that are being used to
improve current data analytics tools are very straightforward mathematical models.

13
1.4 INTRODUCTION TO PYTHON FOR DATA ANALYTICS

Python is a high level, dynamic programming language. Python3.4 version was

used as it is a mature, versatile and robust programming language. It is an
interpreted language which makes the testing and debugging extremely quickly as
there is no compilation step. There are extensive open source libraries available
for this version of python and a large community of users. Python is simple yet
powerful, interpreted and dynamic programming language, which is well known for
its functionality of processing natural language data, i.e. spoken English using
NLTK. Other high level programming languages such as R and Matplotlib were
considered because they have many benefits such as ease of use but they do not
offer the same flexibility and freedom that Python can Deliver.

Features in Python
There are many features in Python, some of which are discussed below
• Easy to code.
• Free and Open Source.
• Object-Oriented Language.
• GUI Programming Support.
• High-Level Language.
• Extensible feature.
• Python is Portable language.
• Python is Integrated language.

1.5 ANACONDA

Anaconda distribution comes with over 250 packages automatically installed, and
over 7,500 additional open-source packages can be installed from PyPI as well as
the conda package and virtual environment manager. It also includes a GUI,
Anaconda Navigator, [12] as a graphical alternative to the Command Line Interface
(CLI).

14
The big difference between conda and the pip package manager is in how package
dependencies are managed, which is a significant challenge for Python data
science and the reason conda exists.
When pip installs a package, it automatically installs any dependent Python
packages without checking if these conflict with previously installed packages. It
will install a package and any of its dependencies regardless of the state of the
existing installation. Because of this, a user with a working installation of, for
example, Google Tensorflow, can find that it stops working having used pip to
install a different package that requires a different version of the dependent numpy
library than the one used by Tensorflow. In some cases, the package may appear
to work but produce different results in detail.
Open source packages can be individually installed from the Anaconda repository,
Anaconda Cloud (anaconda.org), or the user's own private repository or mirror,
using the conda install command. Anaconda, Inc. compiles and builds the
packages available in the Anaconda repository itself, and provides binaries for
Windows 32/64 bit, Linux 64 bit and MacOS 64-bit. Anything available on PyPI
may be installed into a conda environment using pip, and conda will keep track of
what it has installed itself and what pip has installed.
Custom packages can be made using the conda build command, and can be
shared with others by uploading them to Anaconda Cloud, PyPI or other
repositories.
The default installation of Anaconda2 includes Python 2.7 and Anaconda3 includes
Python 3.7. However, it is possible to create new environments that include any
version of Python packaged with conda.

1.6 ANACONDA NAVIGATOR

Anaconda Navigator is a desktop Graphical User Interface (GUI) included in

Anaconda distribution that allows users to launch applications and manage conda
packages, environments and channels without using command-line commands.
Navigator can search for packages on Anaconda Cloud or in a local Anaconda

15
Repository, install them in an environment, run the packages and update them. It
is available for Windows, macOS and Linux.

The following applications are available by default in Navigator:

• JupyterLab.
• Jupyter Notebook.
• QtConsole.
• Spyder.
• Glue.
• Orange.
• RStudio.
• Visual Studio Code.

1.7 JUPYTER NOTEBOOK

Jupyter Notebook (formerly IPython Notebooks) is a web-based interactive

computational environment for creating Jupyter notebook documents. The
"notebook" term can colloquially make reference to many different entities, mainly
the Jupyter web application, Jupyter Python web server, or Jupyter document
format depending on context. Usually ending with the ".ipynb" extension. Jupyter
Notebook can connect to many kernels to allow programming in different
languages. By default, Jupyter Notebook ships with the IPython kernel. As of the
2.3 release[11][12] (October 2014), there are currently 49 Jupyter-compatible
kernels for many programming languages, including Python, R, Julia and Haskell.
The Notebook interface was added to IPython in the 0.12 release[14] (December
2011), renamed to Jupyter notebook in 2015 (IPython 4.0 – Jupyter 1.0). Jupyter
Notebook is similar to the notebook interface of other programs such as Maple,
Mathematica, and SageMath, a computational interface style that originated with
Mathematica in the 1980s. According to The Atlantic, Jupyter interest overtook the
popularity of the Mathematica notebook interface in early 2018.
Pandas is an open-source Python Library providing high-performance data

16
manipulation and analysis tool using its powerful data structures. The name
Pandas is derived from the word Panel Data – an Econometrics from
Multidimensional data. In 2008, developer Wes McKinney started developing
pandas when in need of high performance, flexible tool for analysis of data. Prior
to Pandas, Python was majorly used for data munging and preparation. It had very
less contribution towards data analysis. Pandas solved this problem. Using
Pandas, we can accomplish five typical steps in the processing and analysis of
data, regardless of the origin of data—load, prepare, manipulate, model and
analyze. Python with Pandas is used in a wide range of fields including academic
and commercial domains including finance, economics, Statistics, analytics, etc.

Key Features of Pandas

• Fast and efficient Data Frame object with default and customized
indexing.

• Tools for loading data into in-memory data objects from different
file formats.
• Data alignment and integrated handling of missing data.
• Reshaping and pivoting of date sets.
• Label-based slicing, indexing and sub-setting of large data sets.
• Columns from a data structure can be deleted or inserted.
• Group by data for aggregation and transformations.
• High performance merging and joining of data.
• Time Series functionality.

Matplotlib is the most popular python plotting library. It is a low-level library with a
Matlab like interface which offers lots of freedom at the cost of having to write more
code.

Key features of Matplotlib

17
 • Semantic way to generate complex, subplot grids.

• Settings the aspect ratio of the axes box.

• Colored labels in legends.

• Ticks and labels.

• RcParams can be passed as Decorators.

• 3D plots now support minor ticks.

Plotly allows users to import, copy and paste, or stream data to be analyzed and
visualized. For analysis and styling graphs, Plotly offers a Python sandbox
(NumPy supported), datagrid, and GUI. Python scripts can be saved, shared, and
collaboratively edited in Plotly.

Key features of Plotly

• Charts.

• Dashboards.

• File Export.

• App Manager.

• Kubernetes Authentication.

• Jobs Queue.

• Snapshot Engine.

18
 CHAPTER 2
LITRATURE SURVEY

Birjais et al.[8] experimented on PIMA Indian Diabetes (PID) data set. It has 768
instances and 8 attributes and is available in the UCI machine learning repository.
They aimed to focus more on diabetes diagnosis, which, according to the World
Health Organization (WHO) in 2014, is one of the world’s fastest-growing chronic
diseases. Gradient boosting, logistic regression, and naive Bayes classifiers were
used to predict whether a person is diabetic or not, with gradient boosting having
an accuracy of 86%, logistic regression having a 79% accuracy, and naive Bayes
having a 77% accuracy.

Sadhu, A. and Jadli A.[9] experimented on a diabetes data set taken from the UCI
repository. There were 520 occurrences and 16 attributes in all. They attempted to
concentrate their efforts on predicting diabetes at an early stage. On the validation
set of the employed data set, seven classification techniques were implemented:
k-NN, logistic regression, SVM, naive Bayes, decision tree, random forests, and
multilayer perceptron. The random forests classifier proved to be the best model
for the concerned data set, with an accuracy score of 98%, followed by logistic
regression at 93%, SVM at 94%, naive Bayes at 91%, decision tree at 94%,
random forests at 98%, and multilayer perceptron at 98%, according to the results
of training several machine learning models.

Xue et al.[10] experimented on the diabetes data set taken from the UCI repository;
there were 520 patients and 17 qualities in it. They attempted to concentrate on
early detection of diabetes. They trained on the actual data of 520 diabetic patients
and probable diabetic patients aged 16–90 using supervised ML techniques such
as SVM, naive Bayes classifiers, and LightGBM. The performance of the SVM is
the best when comparing classification and recognition accuracy. The naive
Bayes classifier is the most widely used classification algorithm, with
an accuracy of 93.27%. SVM has the highest accuracy rate of 96.54%. LightGBM

19
has an accuracy of only 88.46%. This demonstrates that SVM is the
best classification algorithm for diabetes prediction.

Le et al.[11] experimented on the early-stage diabetes risk prediction; the data set
used in this research was taken from the UCI repository and consisted of 520
patients and 16 variables. They suggested a ML approach for predicting diabetes
patients’ early onset. It was a new wrapper-based feature selection method that
employed grey wolf optimizer (GWO) and adaptive particle swarm optimization
(APSO) to optimize the multilayer perceptron (MLP) and reduce the number of
needed input attributes. They also compared the results obtained with this method
to those obtained via a variety of traditional machine learning algorithms, including
SVM, DT, k-NN, naive Bayes classifier (NBC), random forest classifier (RFC), and
logistic regression (LR). LR achieved a 95% accuracy rate. k-NN had a
96% accuracy rate, SVM a 95% accuracy rate, NBC a 93% accuracy rate, DT a
95% accuracy rate, and RFC had a 96% accuracy rate. The suggested methods’
computational findings show that not only are fewer features required but also that
higher prediction accuracy may be attained (96% for GWO–MLP and 97% for
APSO–MLP). This research has the potential to be applied in clinical practice and
used as a tool to assist doctors and physicians.

Julius et al.[12] used the Waikato Environment for Knowledge Analysis (Weka)
application platform to test a data set collected from the UCI repository. There were
520 samples in the data set, each with a collection of 17 attributes. The goal of this
study was to use machine learning classification approaches based on observable
sample attributes to predict diabetes at an early stage. The k-NN, SVM, functional
tree (FT), and RFCs were employed as classifiers. k-NN had the
highest accuracy of 98%, followed by SVM at 94%, FT at 93%, and RF at 97%.

Shafi et al.[13] reported that because diabetes is a serious illness, early detection
is always a struggle. This study used machine learning classification methods to
develop a model that could solve any problem and that could be used to identify

20
diabetes development early on. The authors of this research made concerted
efforts to develop a framework that could accurately predict the likelihood of
diabetes in patients. As part of this study, the three ML
approach classification algorithms—DT, SVM, and NBC—were studied and
assessed on various measures. In the study, the PID data set acquired from the
UCI repository was used to save time and produce precise findings. The
experimental results suggested that the NBC approach was adequate, with a
74% accuracy, followed by SVM with a 63% accuracy and the DT with a
72% accuracy. In the future, the built framework, as well as the ML classifiers used,
could be used to identify or diagnose other diseases. The study, as well as several
other ML methodologies, could be extended and improved for diabetes research,
and the scientists intended to classify other algorithms with missing data.

Khanam et al.[14] experimented with diabetes illness prediction. Diabetes is a

condition with no known cure; therefore early detection is essential. In this study,
data mining, ML techniques, and neural network (NN) methodologies were utilized
to predict diabetes. They developed a technique that could accurately predict
diabetes. They used data from the UCI repository’s PID data set. The data set
included information on 768 patients and their 9 attributes. On the data set, they
utilized seven ML methods to predict diabetes: DT, k-NN, RFC, NBC, AB, LR, and
SVM. They used the Weka tool to preprocess the data. They discovered that a
model combining LR and SVM is effective at predicting diabetes. They created a
NN model with two hidden layers and varied epochs and found that the NN with
two hidden layers gave 88.6% accuracy. ANN scored 88.57%, LR scored 78.85%,
NBC scored 78.28%, and RFC scored 77.34%.

Sisodia et al.[15] used the PID data set available on the UCI repository. This data
set contained 768 patients and 8 attributes. They employed three ML
classifications to identify diabetic patients: DT, SVM, and NBC. NBC had the
highest accuracy (76.30%) when compared to the other models.

21
Agarwal et al.[16] used the PID data set of 738 patients as well in their study. To
analyze the effectiveness of this data set for identifying diabetic patients, the
authors applied models such as SVM, k-NN, NBC, ID3, C4.5, and CART. The SVM
and LDA algorithms were the most accurate, with an accuracy of 88%.

Rathore et al.[17] employed classification techniques like SVM and DTs to

predict diabetes mellitus. The PID data set provided the data for this investigation.
PIMA India prioritizes women’s health. The SVM has an accuracy of 82%.
To predict diabetes mellitus, Hassan et al.[18] employed classification approaches
such as the DT, k-NN, and SVM. The SVM outperformed the DT and KNN methods
with a maximum accuracy of 90.23%.

Kandhasamy and Balamurali[19] investigated the prediction accuracy of J48, k-NN,

RFC, and SVM on the diabetes data set. Before preprocessing the data, the author
discovered that the J48 method had a higher accuracy than others, at 73.82%.
After preprocessing, k-NN and RFC demonstrated improved accuracy.

Meng et al.[20] examined J48, LR, and k-NN algorithms on the diabetes data set.
J48 was found to be the most accurate, with a classification accuracy of 78.27%.

Nai-Arun and Moungmai[21] created a web application based on the

prediction accuracy for diabetes prediction. They compared prediction methods
such as DTs, NNs, LR, NBC, and RFC, as well as, bagging and boosting. They
discovered that RFC performed best in terms of accuracy and ROC score, with
an accuracy of 85.558% and an ROC value of 0.912.

Saravananathan and Velmurugan[22] looked at J48, CART, SVM, and k-NN on a

medical data set in their research. They compared them based on accuracy,
specificity, sensitivity, precision, and error rate. With a score of 67.15%, they
discovered that J48 algorithms were the most accurate, followed by SVM
(65.04%), CART (62.28%), and k-NN (53.39%).

22
Kumari and Chitra[23] used SVM, RFC, DT, MLP, and LR, as well as four k-fold
cross-validations (k = 2,4,5,10) in their research. According to the researchers,
MLP with four-fold cross-validation achieves the best accuracy, at 78.7%. They
discovered that MLP outscored all other algorithms.

To predict diabetes, Kavakiotis et al.[24] employed NBC, RFC, k-NN, SVM, DT, and
LR methods. The algorithms were applied using a ten-fold cross-validation
technique. SVM had the best accuracy of all the approaches, measuring 84%,
according to the study.

The work on the classification of “Diabetes Prediction” based on eight attributes

was done by Rawat et al.[25] In this study, five ML algorithms for the analysis and
prediction of diabetic patients were described: AdaBoost, LogicBoost,
RobustBoost, naive Bayes, and bagging. A group of diabetic PIMA Indians was
used to test the proposed strategies. The computed results were found to be quite
accurate, with a classification accuracy of 81.77% and 79.69% for the bagging and
AdaBoost techniques, respectively. As a result, the proposed DM prediction
algorithms were particularly appealing, effective, and efficient.

Perveen et al.[26] used a data set from the Canadian Primary Care Sentinel
Surveillance Network (CPCSSN) database to do their research. The study
employed the AdaBoost and bagging ensemble techniques using the J48 (C4.5)
DT as a base learner and standalone data mining methodology J48 to categorize
patients with diabetes mellitus based on diabetes risk indicators. This
categorization was done across three separate ordinal adult groups in the
CPCSSN. In terms of overall performance, the AdaBoost ensemble method
surpassed both bagging and a single J48 DT, according to the findings.

Mujumdar and Vaidehi[27] presented a diabetes prediction model for better

diabetes classification that included a few extrinsic factors that caused diabetes,

23
as well as regular components such as glucose, BMI, age, insulin, and so on. The
new data set enhanced classification accuracy when compared to the old data set.
Multiple ML approaches were used on the data set, and classification was done
with a variety of algorithms, with LR yielding the highest accuracy at 96%. The
AdaBoost classifier was found to be the most accurate, with a
98.8% accuracy rate. They used two separate data sets to compare
the accuracy of ML techniques. When compared to the existing data set, it was
clear that the model improved diabetes prediction accuracy and precision.

Mercaldo et al.[28] offered a strategy for classifying diabetic patients based on a set
of features chosen according to the WHO criteria. Evaluating real-world data using
state of the art machine learning algorithms. The model was trained using six
alternative classification approaches, with the Hoeffding Tree method scoring
0.770 in precision and 0.775 in recall. They used data from the PIMA Indian
community in Phoenix, Arizona, to evaluate the method.

24
 CHAPTER 3

AIM AND SCOPE OF THE PRESENT INVESTIGATION

3.1 AIM: To accurately predict whether or not the patients in the dataset have
diabetes or not.

3.2 SCOPE: The accuracy of the model is obtained by comparing the predicted
values against the original set of values. Identifying diabetics or predicting the
upcoming of a diabetic life can be propelled by using various machine learning
techniques.

In this project we are going to use a dataset named Diabetes.csv file which
contains the details such as, Pregnancies, Glucose, BloodPressure,
SkinThickness, Insulin, BMI, DiabetesPedigreeFunction, Age and Outcome.

3.3 WHAT IS DIABETES

Diabetes is a condition that is brought on by having an abnormally high level of

blood glucose in the body. Human bodies are in constant need of power, and sugar
is one of the primary sources of vitality that is used in the construction of our
muscles and other tissues. In individuals, the primary reasons of type 2 diabetes
are often an unhealthy habit combined with a lack of physical activity. Diabetes is
a condition that is brought on by an abnormally high level of glucose in the
bloodstream. Diabetes occurs when the pancreas is failing to turn the meal into
insulin; as a result, sugar is not taken into the body, leading to the condition.
Diabetes may cause problems in a variety of body systems, including the kidneys,
eyes, neurological system, arteries, and so on.

OBJECTIVE: The main objective of dataset is to classify that whether a particular

person has diabetes or not with help of several independent variables given in

25
dataset.

Details about the dataset:

The datasets consists of several medical predictor variables and one target
variable, Outcome. Predictor variables includes the number of pregnancies the
patient has had, their BMI, insulin level, age, and so on.

⚫ Pregnancies: Number of times pregnant

⚫ Glucose: Plasma glucose concentration a 2 hours in an oral glucose
tolerance test
⚫ BloodPressure: Diastolic blood pressure (mm Hg)
⚫ SkinThickness: Triceps skin fold thickness (mm)
⚫ Insulin: 2-Hour serum insulin (mu U/ml)
⚫ BMI: Body mass index (weight in kg/(height in m)^2)
⚫ DiabetesPedigreeFunction: Diabetes pedigree function
⚫ Age: Age (years)
⚫ Outcome: Class variable (0 or 1)

26
 CHAPTER 4

EXPERIMENTAL OR REQUIREMENTS AND METHODS

4.1 SYSTEM SPECIFICATION

Hardware Requirements
1. Processor – Intel Core processors or any AMD chips.
2. RAM – 1 GB
3. Hard Disk – 40GB
4. Operating System – Windows 7 and above
Software requirements
1. Jupyter notebook (anaconda 3) or Google collab.
2. Python 3 or latest version.

4.2 PROJECT DESCRIPTION

We are going to predict diadetes via three different supervised machine learning
methods including:SVM,Logistic regression,KNN. We have got this dataset from
Kaggle. We are going to analyze the information using Logistic Regression in this
dataset and try to present in a visual manner using different libraries in python
programming language like numpy,Pandas and matplotlib.

4.3 LIBRARIES USED

We will generally be working on this project using the standard libraries such as:
● Sklearn
● Pandas
● Numpy
● Matplotlib

Sklearn is the root important library from which we are going to import and
make use of the algorithms or the metrics or the train_test_split for making
predictions using the model generated using thee.

27
 Pandas is the standard library one would be dealing with while doing a
machine learning project. As I have mentioned above that I will be using the data
set which is stored in an excel file, we can use pandas to read the excel file.

Similarly we will use Numpy to make use of the numpy arrays, while we are
calculating the errors or handling the data of some kind.

Matplotlib is one of the major necessities as we are going to use it for

creating plots and to analyze our predictions based on that.

We will break down more about the libraries and their usage once we start using
them. I will try to describe that particular usage and its overall description too once
we get into the coding part. As we have seen that the above are the necessary
libraries which we will import as full but for some we will only import the necessary
modules (just to have a quick knowledge of exactly what modules we will be using).

We will import a few specific functions and algorithms from modules of Sklearn
library, they are namely:
● train_test_split from model_selection module of sklearn
● from sklearn.linear_model import LogisticRegression
● metrics module of sklearn

Model_selection is a method for setting a blueprint to analyze data and then

using it to measure new data. Selecting a proper model allows you to generate
accurate results when making a prediction.
train_test_split is a function in sklearn model selection for splitting data
arrays into two subsets: for training data and for testing data.
With this function you don’t need to divide the dataset manually. By default,
sklearn train_test_split will make random partitions for the two subsets. However,
you can also specify a random state for the operation.

28
4.4 Logistic Regression

The predictive analysis which is used for the dependent variable is categorical
called as Logistical Regression. Logistical Regression explains the relationship
between one dependent variable and one or more independent variables.

System Architecture/ Ideation Map:

Fig 4.1: System Architecture

29
Project Implementation:

Steps followed during the implementation

Fig 4.2: Project Implementation

30
Step1:
Went on a search for a diabetes dataset with all the labels given.

Step2:
Then made a flow char of how to implement the diabetes predicting system and
made some conclusions like ,What are the ML algorithms or techniques that would
perfectly fit the problem with good accuracy.

Step3:
Next installation of required software's like Anaconda ,Jupyter, Python.

Step4:
After installing the required software’s ,started with Anaconda lunched Jupyter
notebook with python 3.

Step5:
Now importing all the necessary libraries like Pandas,Numpy,Sklearn,Matplotlib
and all the required modules.

Fig 4.3: code of importing libraries

Step6:
Import the diabetes Dataset into Jupyter notebook.

31
Step7:
Now we need to see the structure of the data like how many columns and rows are
present.

Fig 4.4: code for importing data set

Step8:
Then started looking for missing values and filling those with mean of the particular
column (here we have 2 missing values one in SkinThickness and the other on
Insulin)

32
 Fig 4.5: Example code for the pre-work of data set with null values

Fig 4.6: Example code for the pre-work of data set with null values

Step9:
Once we are good with the data with no null values and missing values and no
string values then we can split the data into 2 parts.
Part1 : Training data
Part2: Testing data

33
Here our Target column is ‘Outcome’ i.e y so separating from the main data now
we have y data frame with ‘Outcome’ and x data frame with all the other lables.
We need at least 80% of data to train and 20% of data to test.

Fig 4.7 : code for splitting data frame into two parts based on the column
values

Step10:
Now further splitting the data into
x_train,x_test
y_train,y_test

34
Fig 4.8: code for splitting data frame into two parts based on the column
values

Fig 4.9: Splitting Datasets

35
Step11:
Now we will apply ML algorithm i.e. LogisticRegression from Sklearn.linear_model.

Fig 4.10: code to demonstrate fitting the algorithm to the model and
resulting array of predictions made

Step12:
From the trained model we will calculate the accuracy of our model.

36
 Fig 4.11: code to calculate the accuracy of the model

Step13:
Now will see the accuracy graph.

Fig 4.12: Accuracy graph model

37
Confusion Matrix

Fig 4.13: code for importing metrics to find the accuracy score

METHODOLOGY:

⚫ Data Collection
⚫ Data Preparation
⚫ Choosing a model
⚫ Training the model
⚫ Evaluating the model
⚫ Parameter turning
⚫ Making prediction

◼ Data Collection : Collecting the Dataset from the Kaggle.com

◼ Data Preparation: We need to prepare data before training so preparation
includes dealing with missing data filling the null values with the mean or
median.
◼ Choosing a Model: model or algorithm selection is one of the most important
step in predictions which is to select a particular ML Algorithm for training.

38
◼ Training the Model: After selecting the model we need to train the data.
◼ Evaluating the Model: After training we need to check weather the model is
accurately predicting the outcomes(Accuracy Score=75%).
◼ Making Predictions: Finally we need to predict the diabetes of a particular row
to check weather the model is working perfectly or not.

TECHOLOGIES USED:

◼ Data science is a multidisciplinary approach to extracting actionable insights

from the large and ever-increasing volumes of data collected and created by
today’s organizations.
◼ Data science encompasses preparing data for analysis and processing,
performing advanced data analysis, and presenting the results to reveal
patterns and enable stakeholders to draw informed conclusions.
◼ Data preparation can involve cleansing, aggregating, and manipulating it to be
ready for specific types of processing. Analysis requires the development and
use of algorithms, analytics and AI models.
◼ It’s driven by software that combs through data to find patterns within to
transform these patterns into predictions that support business decision-
making.
◼ The accuracy of these predictions must be validated through scientifically
designed tests and experiments.
◼ The results should be shared through the skillful use of data visualization tools
that make it possible for anyone to see the patterns and understand trends.

39
 CHAPTER 5
RESULT, PERFORMANCE ANALYSIS

Fig 5.1: code for the prediction model

◼ From the above snippet we are successfully able to see the predictions made
by the model.
◼ If the Output is ‘1’ that means the person is diabetic but if the Output is ‘0’ that
means the person is not diabetic.
◼ Here the Output is based on considering the other labels like Glucose,
Pregnancies , BloodPressure , SkinThickness , Insulin , BMI ,
DiabetesPedigreeFunction , Age.
◼ Successfully achieved an accuracy of 75%

40
 Code to calculate the accuracy of the model

◼ Successfully achieved an accuracy of 75% using LogisticRegression and 69%

accuracy using DecisionTreeClassifier.

41
 CHAPTER 6
SUMMARY AND CONCLUSION

The main aim of this project was to design and implement Diabetes Prediction
Using Machine Learning Methods and Performance Analysis of that methods
and it has been achieved successfully. Successfully able to clean the data
and split it into training and testing data. The proposed approach uses various
classification and ensemble learning method in which Decision Tree, Logistic
Regression are used. 75% classification accuracy has been achieved. The
Experimental results can be asst health care to take early prediction and make
early decision to cure diabetes and save humans life. In future I would like to
move on to Deep Learning and upgrading myself with new technologies like
TensorFlow and keras and NeuralNetwork and continue my research in the
field of AI and built many more projects.

42
APPENDIX

A. Source Code

{#c1}
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from sklearn.model_selection import train_test_split
from sklearn.tree import DecisionTreeClassifier
from sklearn import metrics
from sklearn.linear_model import LogisticRegression

{#c2}
data=pd.read_csv("diabetes.csv")
data

{#c3}
data.isna().sum()

{#c3}
mean_SkinThickness=data.SkinThickness.mean()

{#c4}
mean_Insulin=data.Insulin.mean()

{#c5}
data['SkinThickness'].fillna(value=mean_SkinThickness,inplace=True
)
data

{#c6}
x=data.drop('Outcome',axis = 1)
y=data.Outcome
print(x,y)

{#c7}

43
from sklearn.model_selection import train_test_split
x_train,x_test,y_train,y_test=train_test_split(x,y,test_size=0.20,stratify
=y,random_state=2)

{#c8}
x_train.shape

{#c9}
from sklearn.linear_model import LogisticRegression
model = LogisticRegression()
model.fit(x_train,y_train)
x_test = x_test.fillna(x_train.mean())
y_pred=model.predict(x_test)
y_pred

{#c10}
accuracy=metrics.accuracy_score(y_test,y_pred)
Precision=metrics.precision_score(y_test, y_pred)
Recall=metrics.recall_score(y_test, y_pred)
print("Accuracy",accuracy)
print("precision",Precision)
print("Recall",Recall)

{c11}
y_pred_proba = model.predict_proba(x_test)[::,1]
fpr, tpr, _ = metrics.roc_curve (y_test, y_pred_proba)
auc = metrics.roc_auc_score (y_test, y_pred_proba)
plt.plot(fpr, tpr, label="data 1, auc="+str(auc))
plt.legend(loc=4)
plt.show()

{#c12}
from sklearn import metrics

cnf_matrix = metrics.confusion_matrix(y_test, y_pred)

cnf_matrix

{#c13}
#Predicting system

44
test_sample = (5,166,72,19,175,25.8,0.587,51)
test_sample_as_numpy_array = np.asarray(test_sample)
input_data_reshape = test_sample_as_numpy_array.reshape(1,-1)
prediction = model.predict(input_data_reshape)
print(prediction)

if (prediction[0]== 0):
print('The person is not diabetic')
else:
print('The person is diabetic')

45
B. Screen Shots

Fig 6.1: code of importing libraries

46
 Fig 6.2: code for importing data set

Fig 6.3: Example code for the pre-work of data set with null values

Fig 6.4: Example code for the pre-work of data set with null values

47
Fig 6.5: code for splitting data frame into two parts based on the column
values

48
Fig 6.6: code for splitting data frame into two parts based on the column
values

Fig 6.7: code to demonstrate fitting the algorithm to the model and
resulting array of predictions made

49
Fig 6.8: code to calculate the accuracy of the model

Fig 6.9: Accuracy graph model

50
Fig 6.10: code for importing metrics to find the accuracy score

Fig 6.11: code for the prediction model

51
REFERENCES

1. DeeptiSisodia, Dilip Singh Sisodia,”Prediction of Diabetes Using

Classification Algorithm”, www.elsevier.com/locate/procedia,
Procedia computer science 132(2018) 1578-1585.

2. Xue-Hui Meng,Yi-Xiang Huang,Dong-PingRao,Qiug

Liu,2013,”Comparison of Three Data Mining Models For Predicting
Diabetes of Prediabetes By Rick Factos”,Kaohsiung journal of
medical science(2013) 29,93-99.

3. V.AnujaKumari, R.Chithra.”Classification of Diabetes Disease Using

Support Vector Machine”.vol 3.,Issue 2,March-April 2013,pp.1797-
1801.www.ijera.com.

4. Monisha.A, S.ShalinChistina, Nirmala Santiago, “Decision support

system for a chronic disease-Diabetes”. International Journal of
Computer &Mathematical Science(IJCMS),ISSN 2347-8527 Volume
7,Issue 3,March 2018.

5. S.Selvakumar, K.Senthamarai Kannan and S.GothaiNachiyar,

“Prediction of Diabetes Diagnosis Using Classification Based Data
Mining Techniques”, International Journal of statistics and Systems,ISSN
0973-2675 Volume 12,Number 2(2017),PP.183-
188.http://www.ripublication.com.

6. Aiswarya Iyar, S. Jeyalatha and RonakSumbaly, “Diagnosis of

Diabetes Using Classification Mining Techniques”, International
Journal of Data Mining & Knowledge Management Process (IJDKP)
Vol.5, No.1, January 2015.

7. B.Tamilvanan, Dr.V.MuraliBhaskaran, “An Experimental Study of

Diabetes Disease Prediction System Using Classification Techniques”,
IOSR Journal of Computer Engineering (IOSR-JCE), e-ISSN: 2278-

52
 0661,p-ISSN: 2278-8727, Volume 19, Issue 1, Ver. IV (Jan.-Feb. 2017),
PP 39-44, www.iosrjournals.org.

8. Rahul Joshi, MinyechilAlehegn,“Analysis and prediction of diabetes

diseases using machine learning algorithm: Ensemble
approach”,International Research Journal of Engineering and
Technology (IRJET),Volume: 04 Issue: 10 | Oct -2017,e-ISSN: 2395-
0056,p-ISSN: 2395-0072. www.irjet.net.

9. Amina Azar,Yasir Ali, Muhammad Awais, KhurramZaheer,“Data Mining

Models Comparison for Diabetes Prediction”, (IJACSA) International
Journal of Advanced Computer Science and Applications, Vol. 9, No. 8,
2018.

10. VeenaVijayan.V, Anjali.C,”Decision Support Systems for Predicting

Diabetes Mellitus –A Review”, Proceedings of 2015 Global Conference
on Communication Technologies(GCCT 2015), 978-1-4799-8553-
1/15/$31.00 © s2015 IEEE.

11. DeepikaVerma , Dr.Nidhi Mishra, “Analysis and prediction of breast

cancer and diabetes disease dataset using data mining classification
techniques”,Proceedings of the International Conference on Intelligent
Sustainable Systems (ICISS 2017),IEEE Xplore Compliant - Part
Number:CFP17M19-ART, ISBN:978-1-5386-1959-9.

12. VeenaVijayan V, Anjali c, “Prediction of diagnosis of diabetes mellitus

–A machine learning approach ”, 2015 IEEE Recent Advances in
Intelligent Computational Systems (RAICS) | 10-12 December 2015.

13. Ayman Mir, SudhirN.Dhage, “Diabetes Disease Prediction using

MachineLearning on Big Data of Healthcare”, 2018 Fourth
International Conference on Computing Communication Control and
Automation (ICCUBEA).

14. S M Hasan Mahmud, MdAltabHossin, Md. Razu Ahmed,

53
 SheakRashedHaiderNoori, MdNazirul Islam Sarkar, “Machine Learning
Based Unified Framework for DiabetesPrediction”, BDET 2018, August
25–27, 2018, Chengdu, China. © 2018 Association for Computing
Machinery. ACM ISBN 978-1-4503-6582-
6/18/08…$15.00.DOI:https://doi.org/10.1145/3297730.3297737.

15. AakanshaRathore, Simran Chauhan, SakshiGujral, “Detecting and

Predicting Diabetes Using Supervised Learning: An Approach towards
Better Healthcare for Women”,Volume 8, No. 5, May-June 2017,
ISSN No. 0976-5697, Available Online at www.ijarcs.info.

54