This package uses machine learning models to predict patient outcomes based on various factors such as medical history, treatment plans, and demographic data. It helps in identifying high-risk patients and optimizing treatment strategies.
pip install patient_outcome_predictorDescription: The dataset used in this demo contains simulated medical records for a fictional group of patients. The dataset was generated using the Python Faker library to create realistic but fake data. The dataset includes the following fields for each patient:
Patient ID: A unique identifier for each patient (integer). Name: A randomly generated full name (string). Date of birth: A randomly generated date of birth with ages between 1 and 100 years old (date). Gender: A randomly selected gender (M or F) (string). Medical conditions: A list of three random, unique words representing medical conditions (string). Medications: A list of three random, unique words representing medications (string). Allergies: A list of three random, unique words representing allergies (string). Last appointment date: A randomly generated date within the range of the last 2 years (date). Please note that this dataset is for demonstration and testing purposes only. The data is entirely fictional and should not be used for any decision-making or analysis.
This enhanced implementation includes real-time prediction, CSV preprocessing, and a Streamlit app for demonstration. The fictional dataset provides a realistic context for testing the package.
#Example Usage
import pandas as pd
from patient_outcome_predictor import DataPreprocessor, ModelTrainer, OutcomePredictor
# this loads dataset
df = pd.read_csv('dataset.csv')
# Preprocess the data
preprocessor = DataPreprocessor()
X_train, X_test, X_val, y_train, y_test, y_val = preprocessor.preprocess(df, 'outcome')
# Train the model
model_trainer = ModelTrainer()
model_trainer.train(X_train, y_train)
# Evaluation of the model
accuracy = model_trainer.evaluate(X_test, y_test)
print(f'Accuracy: {accuracy}')
# Predict outcomes
predictor = OutcomePredictor(model_trainer.get_model(), preprocessor.scaler)
predictions = predictor.predict(X_test)
print(predictions)