Stroke Prediction: EDA and Accuracy
A data analysis and machine learning project to predict stroke occurrences and achieve high accuracy. This project includes exploratory data analysis (EDA) to gain insights from the dataset and build machine learning models for stroke prediction.
Table of Contents
- Overview
- Dataset
- Exploratory Data Analysis (EDA)
- Machine Learning Models
- Getting Started
- Contributing
Overview
Stroke is a serious medical condition that can have severe consequences. This project aims to use data analysis and machine learning to predict the likelihood of a stroke occurrence. By analyzing a dataset containing various health-related features, we gain insights and create models for accurate stroke prediction.
Dataset
1)id: Identification number of the individual.
2)gender: Gender of the individual.
3)hypertension: Health related parameter, does person have hypertension.
4)heart_disease: Health related parameter, does person have heart disease.
5)ever_married: Personal information, is person married on not?
6)work_type: Nature of work place.
7)Residence_type: Residence type of the individual.
8)avg_glucose_level: average glucose level in blood for the individual.
9)bmi: body mass index of the individual.
10)smoking_status: Habitual information. Current smoking status of individual.
11)stroke: Our taget, is person suffered heart attack?
Exploratory Data Analysis (EDA)
In the EDA phase, we explore and visualize the dataset to understand its characteristics. EDA helps us identify trends, correlations, and anomalies in the data. It involves data cleaning, visualization, and statistical analysis.
Machine Learning Models
Using machine learning, we create predictive models based on the dataset. These models are trained to predict the Gradient Boosting Algorithm of a stroke occurrence. The project includes various machine learning algorithms, such as logistic regression, decision trees, and random forests.
Getting Started
Run the Jupyter Notebook containing the EDA and machine learning code.
Explore the project, run the code, and analyze the results.
Contributing
Contributions to this project are welcome.