Flu-vaccine-predictor

Description:

This project aims to predict the likelihood of individuals receiving the xyz and seasonal flu vaccines. Utilizing a comprehensive dataset with various demographic, behavioral, and opinion-based features, this model estimates the probabilities for each type of vaccine uptake. The primary objectives are to accurately forecast two probabilities: one for the xyz vaccine and one for the seasonal flu vaccine.

Project Features:

• Multilabel Classification: The problem is formulated as a multilabel classification, where each individual may receive either, both, or neither of the vaccines.

• Data Preprocessing:

  • Handling missing values by filling them with the mode.
  • Encoding categorical variables using one-hot encoding to convert them into numerical format.
  • Ensuring consistent feature alignment between training and test datasets.

• Model Training:

  • Employing a RandomForestClassifier wrapped in a MultiOutputClassifier to handle the multilabel nature of the task.
  • Splitting the data into training and validation sets to evaluate model performance.

• Evaluation Metrics:

  • Using the ROC AUC score to measure the performance of the model for each target variable.
  • Calculating the average ROC AUC score to provide an overall assessment of model performance.

• Prediction:

  • Generating probability predictions for the test dataset.
  • Outputting the predictions in a format suitable for submission, including respondent IDs and their associated vaccine uptake probabilities.

Dataset Description:

The dataset comprises 36 columns:

• respondent_id: A unique identifier for each individual.
• Features: 35 features including demographic details (age, sex, race, etc.), behavioral patterns (mask usage, handwashing, etc.), and opinions on vaccine effectiveness and risks.

Usage:

To use this project:

1. Clone the repository.
2. Ensure you have the required datasets in the specified directory.
3. Run the provided Python script to preprocess the data, train the model, and generate predictions.

Dependencies:

• pandas
• scikit-learn
• notebook (Jupyter)

How to Run:

# Clone the repository
git clone https://github.com/your-username/Flu-vaccine-predictor.git

# Navigate to the project directory
cd Flu_vaccine_predictor

# Ensure the datasets are placed in the specified directory
# Run the script
jupyter notebook Flu_vaccine_predictor.ipynb

This project provides a robust framework for predicting vaccine uptake probabilities using machine learning techniques, ensuring a comprehensive approach to data preprocessing, model training, and evaluation.