{"id":23929418,"url":"https://github.com/munavarhs/patient-readmission-analysis","last_synced_at":"2025-10-29T04:23:37.223Z","repository":{"id":269121537,"uuid":"906493063","full_name":"munavarhs/Patient-Readmission-Analysis","owner":"munavarhs","description":"Hospital readmissions, particularly within 30 days of discharge, are a significant issue in healthcare. This project leverages machine learning to predict the likelihood of readmissions using patient data, enabling healthcare providers to target high-risk patients and improve outcomes.","archived":false,"fork":false,"pushed_at":"2024-12-21T04:07:23.000Z","size":315,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"main","last_synced_at":"2025-01-05T23:17:27.586Z","etag":null,"topics":["classification","decision-tree","ensemble-model","gradient-boosting","logistic-regression","machine-learning","naivebayes","random-forest","xgboost-classifier"],"latest_commit_sha":null,"homepage":"","language":"Jupyter Notebook","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/munavarhs.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2024-12-21T04:03:29.000Z","updated_at":"2024-12-21T04:07:25.000Z","dependencies_parsed_at":"2024-12-21T05:17:29.605Z","dependency_job_id":"3eab61b9-0109-4917-847b-df35e6ade351","html_url":"https://github.com/munavarhs/Patient-Readmission-Analysis","commit_stats":null,"previous_names":["munavarhs/patient-readmission-analysis"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/munavarhs%2FPatient-Readmission-Analysis","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/munavarhs%2FPatient-Readmission-Analysis/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/munavarhs%2FPatient-Readmission-Analysis/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/munavarhs%2FPatient-Readmission-Analysis/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/munavarhs","download_url":"https://codeload.github.com/munavarhs/Patient-Readmission-Analysis/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":240404520,"owners_count":19796059,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["classification","decision-tree","ensemble-model","gradient-boosting","logistic-regression","machine-learning","naivebayes","random-forest","xgboost-classifier"],"created_at":"2025-01-05T23:17:33.788Z","updated_at":"2025-10-06T12:04:47.690Z","avatar_url":"https://github.com/munavarhs.png","language":"Jupyter Notebook","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Hospital Readmission Prediction\n\n## Overview\nHospital readmissions, particularly within 30 days of discharge, are a significant issue in healthcare. This project leverages **machine learning** to predict the likelihood of readmissions using patient data, enabling healthcare providers to target high-risk patients and improve outcomes.\n\n---\n\n## Problem Statement\nHospital readmissions pose challenges in terms of:\n- **Financial Costs**: Increased costs and penalties due to programs like the Hospital Readmissions Reduction Program (HRRP).\n- **Patient Health**: Unresolved health issues leading to poorer outcomes.\n- **Resource Allocation**: Inefficient utilization of hospital resources.\n\nOur solution uses predictive modeling to address these issues by identifying patients likely to be readmitted.\n\n---\n\n## Dataset\nWe utilized a publicly available dataset from [Kaggle](https://www.kaggle.com/datasets/gbiamgaurav/patient-survival-prediction), containing:\n- **Records**: 91,713 patient entries.\n- **Features**: 186 attributes, expanded to 239 after preprocessing.\n- Key features include:\n  - Demographics (age, gender, ethnicity).\n  - Clinical measures (glucose levels, heart rate, BMI).\n  - Admission details (ICU ID, prior visits, readmission status).\n\n---\n\n## Methods\nWe implemented the following **machine learning models** for binary classification:\n1. **Logistic Regression**: Provided a baseline for comparison.\n2. **Random Forest**: Captured complex relationships with hyperparameter tuning.\n3. **Gradient Boosting**: Improved accuracy with iterative optimization.\n4. **XGBoost Classifier**: Handled class imbalance effectively.\n5. **Ensemble Model**: Combined predictions from Random Forest and XGBoost for superior performance.\n6. **Decision Tree**: Used as a benchmark.\n7. **Naive Bayes**: Provided a simpler probabilistic approach.\n\n### Preprocessing Steps\n1. **Data Cleaning**: Removed duplicates and handled missing values.\n2. **Feature Engineering**: Added \"prior visits\" as a new feature.\n3. **Encoding**: Converted categorical variables using one-hot encoding.\n4. **Scaling**: Normalized numerical columns for balanced feature importance.\n\n---\n\n## Results\n| Model                | Accuracy | Precision (Class 1) | Recall (Class 1) | AUC-ROC |\n|----------------------|----------|---------------------|------------------|---------|\n| Logistic Regression  | 91.9%    | 0.54                | 0.63             | 0.87    |\n| Random Forest        | 92.3%    | 0.54                | 0.68             | 0.89    |\n| Gradient Boosting    | 92.7%    | 0.62                | 0.66             | 0.87    |\n| XGBoost Classifier   | 93.0%    | 0.63                | 0.70             | 0.90    |\n| **Ensemble Model**   | **93.3%**| **0.65**            | **0.72**         | **0.91**|\n| Decision Tree        | 90.2%    | 0.52                | 0.61             | 0.86    |\n| Naive Bayes          | 80.0%    | 0.25                | 0.67             | 0.82    |\n\n### Key Observations\n- **Ensemble Model** achieved the best performance with 93.3% accuracy and an AUC-ROC of 0.91.\n- **SHAP analysis** highlighted the most impactful features for predictions, enhancing model interpretability.\n\n---\n\n## Impact\n- **Healthcare Providers**: Improved patient care through targeted interventions.\n- **Hospitals**: Reduced readmission rates, saving costs and resources.\n- **Patients**: Better health outcomes and fewer complications.\n\n---\n\n## Contributions\n- **Munavar Hussain**: Model selection and evaluation.\n- **Sujith Reddy A**: Hyperparameter tuning, data analysis, and report writing.\n- **Irfan Mohammed**: Data preprocessing and feature engineering.\n\n---\n\n## Installation and Usage\n1. Clone the repository:\n   ```bash\n   git clone https://github.com/munavarhs/Patient-Readmission-Analysis/\n   ```\n\n---\n\n## License\nThis project is licensed under the MIT License. See the `LICENSE` file for details.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmunavarhs%2Fpatient-readmission-analysis","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fmunavarhs%2Fpatient-readmission-analysis","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmunavarhs%2Fpatient-readmission-analysis/lists"}