Kidney Disease Prediction: using Ensemble Learning

The primary goal of the Kidney Disease Prediction Model is to assist in identifying potential cases of kidney disease by analyzing a set of input attributes related to a patients health and laboratory test results. The model is developed using an ensemble approach, which combines multiple machine learning algorithms to enhance the accuracy of predictions.

To use the model, users can access the live demo application provided in the repository. They are required to input specific attributes such as age, blood pressure, urine-specific gravity, albumin level, sugar level, red blood cell count, pus cell count, presence of pus cell clumps and bacterial infection, blood glucose level, blood urea level, serum creatinine level, sodium and potassium levels, hemoglobin level, packed cell volume level, white blood cell count, presence of hypertension, diabetes mellitus, coronary artery disease, appetite status, pedal edema, and anemia. The model then processes this data to provide predictions on whether the patient is likely to have kidney disease.

The repository also includes a selection of ensemble learning models that are used in building the prediction model. Users can compare the performance of these individual models with the ensemble model to gain insights into their effectiveness.

• Ensemble Learning: The model leverages ensemble learning techniques, which combine multiple machine learning algorithms to improve predictive accuracy.
• High Accuracy: The kidney disease prediction model has achieved an impressive 100% accuracy on the test dataset, indicating its efficacy in making accurate predictions.
• Input Attributes: Users can input various attributes related to a patients health and laboratory test results, such as age, blood pressure, specific gravity of urine, albumin level, sugar level, and many others.
• Live Demo: The repository includes a live demo application that allows users to interact with the model and obtain predictions based on their input attributes.
• Multiple Models: The model is built using a range of ensemble learning algorithms, including Random Forest, Naïve Bayes, Logistic Regression, Decision Tree, Gradient Boosting, Support Vector Machine, LightGBM, XGBoost, Multilayer Perceptron (MLP), and Artificial Neural Network (ANN).
• Performance Metrics: The application provides performance metrics for each model, including accuracy, precision, recall, and F1 score, allowing users to assess and compare the models.
• Data Source: While the kidney disease dataset used for training and testing the model is not included in the repository, users can find publicly available datasets or use their own dataset following the format described in the Input Attributes section.
• Disclaimer: The repository includes a disclaimer that highlights the models predictions are based on available data and model performance, and users should consult a medical professional for a reliable diagnosis. The authors and contributors are not responsible for any decisions or actions made based on the models predictions.