ahmedburale / healthcare-data-analysis

Healthcare Data Analysis and Predictive Modeling

Project Overview

This project aims to analyze synthetic healthcare data and build a machine learning model to predict test results based on various factors such as age, gender, and medical condition.

Dataset

The dataset contains synthetic healthcare records with the following fields:

• Patient Name
• Age
• Gender
• Blood Type
• Medical Condition
• Date of Admission
• Doctor Name
• Hospital Name
• Insurance Provider
• Billing Amount
• Room Number
• Admission Type
• Discharge Date
• Medication
• Test Results

Steps Involved

1. Data Cleaning: Standardize column names, correct data types, and handle missing values.
2. Exploratory Data Analysis (EDA): Visualize the distribution of age, gender, medical conditions, admission types, and test results.
3. Feature Engineering: Encode categorical variables, create new features such as the length of hospital stay, and drop unnecessary columns.
4. Correlation Analysis: Calculate and visualize the correlation matrix to understand relationships between features.
5. Model Building and Evaluation: Train a Random Forest classifier and evaluate its performance using accuracy, classification report, and confusion matrix. Visualize feature importance.

Getting Started

Prerequisites

• Python 3.x
• Required libraries: pandas, matplotlib, seaborn, scikit-learn

Running the Project

1. Clone the repository.
2. Install the required libraries using pip install -r requirements.txt.
3. Run the healthcare_analysis.ipynb notebook to see the analysis and model building steps.

Results

The analysis provides insights into the data distribution and relationships between features. The Random Forest classifier model predicts test results with a certain accuracy, and the feature importance plot shows which features contribute the most to the predictions.

Conclusion

This project demonstrates the process of cleaning and analyzing healthcare data, performing exploratory data analysis, and building a predictive model using machine learning techniques. '''

Note

This analysis is still under regular update on insights...