You are on page 1of 14

DATA MINING IN HEALTH CARE - PROGNOSTICATE AND DIAGNOSIS OF

DIABETES

First Doctoral Committee Meeting on 19/12/2018

by
K. Kanmani
Asst. professor of Computer Applications
SRM University.

under the guidance of


Dr A. MURUGAN
Associate Professor and Head
PG and Research Department of computer Science
Dr. Ambedkar Govt. Arts College
Vyasarpadi.
DIABETES
 The area of the proposed research work is to predict
the diabetes using various data mining techniques.
 Diabetes can be referred to as a chronic disease
characterized by a high level of sugar (glucose) in
blood.
• Diabetes is one of the main causes of death, disability
and economic loss through out the world.
Types of diabetes
 Type 1 Diabetes Mellitus
 Type 2 Diabetes Mellitus
 Gestational Diabetes
Type1 diabetes
 Insulin-dependent diabetes, is a chronic condition in which
the pancreas produces little or no insulin. Insulin is a
hormone needed to allow sugar (glucose) to enter cells to
produce energy.
 Symptoms
Increased thirst
Frequent urination
Bed-wetting in children who previously didn't wet the
bed during the night
Extreme hunger
Unintended weight loss
Irritability and other mood changes
Fatigue and weakness
Blurred vision
Type2 diabetes
 It is a long-term metabolic disorder that is
characterized by high blood sugar, insulin resistance,
and relative lack of insulin
 genetics
 lack of exercise
 being overweight
Data mining in health care
 knowledge discovery in databases (KDD).
 Data mining is about finding new information in a
lot of data.
• Data mining holds great potential for the healthcare
industry to enable health systems to systematically use
data and analytics to identify inefficiencies and best
practices that improve care and reduce costs
Decision Tree
 A Flow chart like Tree Structure
 Internal nodes denotes a test on an attribute
 Branch represents an outcome of the test
 Leaf notes represents class labels or class distribution
Decision tree – contd
Decision tree construction consist of two phases
Decision Tree Construction
All the training examples are at the root
Partition examples recursively based on selected
attributes.
Tree Pruning
Identify and remove branches that reflect noise
Methodology – Research
Observation
The Study consist of two stages

Data Processing
Attribute Identification and selection
Handling Missing values
Numerical Discretization
Decision Tree Construction
separate data into fixed number of
partitions
Methodology - contd
 Select the first fold for testing, remaining folds are
used for training
 Perform classification and obtain performance metrics
 Select the next partition as testing and use the rest as
training data
 Repeat classification until each partition has been
used as the test set
Data set - Attributes
 Pregnancies Number of times pregnant
 Glucose Plasma glucose concentration a 2 hours in an
oral glucose tolerance test
 Blood Pressure Diastolic blood pressure
 Skin Thickness Triceps skin fold thickness
 Insulin2-Hour serum insulin
 BMI Body mass index (weight in kg/(height in m)^2)
 Age (years)
 Outcome Class variable (0 or 1)
conclusion
The aim of the research work is to investigate the risk
factors associated with the diabetes and to make
accurate predictions
Decision tree is one of the most powerful and widely
applied technique for classification and prediction. It
produces reasonably good models. It gives benefits to
society to make decision regarding the diagnosis of
diabetes.
References
[1] Jiawei and kamber, “Data Mining concepts and techniques”
[2] Maham Jahangir et.al, “An expert system for diabetes prediction using
Auto tuned Multi layer Perceptron”, IEEE, 2017, pp .722-728
[3] Asma A. Aljarullah, “Decision Tree Discovery for the diagnosis of Type
II Diabetes”, IEEE, 2011, pp. 303-307
[4] Nongayo Nai arun, Rungruttikarn Moungmai, “ Comparison of
classifiers for the risk of diabetes prediction”, elsevier, 2015
[5]Haroon kaur, Shalini Batra, “HPCC: An ensembled framework for the
prediction of the onset of diabetes”, IEEE International conference on
signal processing computing and control, 2017.

You might also like