Scribd
Upload
19 views10 pages

Heart Attack Prediction Using Machine Learning

Uploaded by

suchetkapoor0
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
19 views10 pages

Heart Attack Prediction Using Machine Learning

Uploaded by

suchetkapoor0
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 10

HEART ATTACK

PREDICTION
USING MACHINE
LEARNING:A
NOVEL ENSEMBLE
LEARNING
APPROACH

P R E S E N T E D B Y :
S U C H E T K A P O O R
NCIESA-2024.
TABLE OF CONTENT T

Brief introduction about research paper It tells us about what my core objective of paper

Initial Requirements and dataset information About what are intial requirements, where to download dataset.

Data Preprocessing Dataset preprocessing steps for removing null value,categorial value.

Flowchart of proposed algorithm Proposed algorithm flowchart from starting to end.

Algorithm Approach The steps of predicted RBFC algorithm and approach.

Results and outcome Final outcome and performance metrics.


BRIEF INTRODUCTION
ABOUT THE RESEARCH
Heart attack accounts for about 20-25 million deaths worldwide. However, 80% of these deaths are
premature deaths and can be reversed by introducing new age tools in biomedical diagnosis.
Machine learning, one of the new age tools, offers a better, cost-effective approach to minimizing
these deaths. In the past, machine learning could analyze vast amounts of data, including health
records and scans, and identify patterns. This allows researchers to develop a model based on their
results. Utilizing ML In this paper, we proposed a hybrid model named random forest boosting
classifier (RFBC), a combination of random forest and gradient boosting classifiers. The proposed
model achieved an accuracy of 91.8%. The dataset used was the standard Cleveland heart disease
dataset. The results indicated that despite advancements in the medical field, combining Ml with
medical diagnosis techniques remains an easy-to-implement solution.

10.10
Initial Requirements

Kaggle Link : https://www.kaggle.com/code/suchetkapoor/heart-


attack
UCI heart attack dataset link:
https://archive.ics.uci.edu/dataset/45/heart+disease
This study used the Cleveland heart disease

DATASET dataset from the UCI repository. Our dataset


contains a record of 303 patients.

USED The dataset contains fourteen features: age,


sex, cp, threstbps, chol, fbs, restecg, thalach,
exang, old peak, slope, ca, and thal.

04.10
DATA
PREPROCESSING
Data preprocessing steps:

1. Check for null values: Removed due to potential


inaccuracies and error rates in algorithms.
2. Identify outliers: Removed inconsistencies for improved
data quality and accuracy.
3. Eliminate duplicate values: Removed redundant
records for better reliability and efficiency.
4. Analyze data distribution: For insights, examine the
balance between normal and abnormal patient records.

Preprocessing results: Dataset is free from null values,


outliers, and duplicate values, ensuring data integrity and
accuracy.
PROPOSED ALGORITHM FLOWCHART
1. Preprocessing: The data is then
preprocessed to handle categorical, null,
and duplicate variables.
2. Splitting the data: The data is split into
training and testing sets (typically 80%
training and 20% testing).
3. Hybrid ensemble model: A hybrid
ensemble model, possibly Random
Forest and Gradient Boosting Classifier,
is applied to the training data.
4. Predicted output: The model generates
predictions on the testing data.
5. Performance evaluation: The model’s
performance is evaluated using metrics.
ALGORITHM
Steps APPROACH
1. Base Model Training: Train the base
models, RF with 40 trees and GB with 30
estimators and a learning rate of 0.001,
on the Cleveland dataset.
2. Ensemble Model Creation: Create a VC
ensemble model with the RF and GB
models as estimators, using 'soft' voting
based on class probabilities.
3. Model Fitting: Fit the hybrid model
using the training data.
4. Prediction: Utilize the trained hybrid
model to predict the test data and obtain
predicted values.
5. Performance Evaluation: Evaluate the
model performance on the test data using
appropriate metrics such as accuracy,
precision, recall, and F1-score.
OUR RESULTS
The diagram shows that our hybrid model RFBC
outperformed other machine learning models in
terms of accuracy precision and recall.

RFBC

ACCURACY PRECESION RECALL

91.8 92 92

Table1. THE PROPOSED MODEL PERFFORMANCE


CONCLUSION & FUTURE WORK
Implementing embedded AI and LLMs (Learning
Management Systems) to enhance heart attack
prediction in India, providing personalized and real-
time monitoring for individuals.

Applying deep learning techniques, such as


Convolutional Neural Networks (CNNs), on medical
images to improve prediction accuracy and reduce
time-to-diagnosis.

Collaborating with healthcare professionals and


organizations to ensure practical implementation and
continuous improvement of AI-driven heart attack
prediction models.

N
NCIESA-2024

You might also like