Summary of Comparative Analysis Of Lstm Neural Networks and Traditional Machine Learning Models For Predicting Diabetes Patient Readmission, by Abolfazl Zarghani
Comparative Analysis of LSTM Neural Networks and Traditional Machine Learning Models for Predicting Diabetes Patient Readmission
by Abolfazl Zarghani
First submitted to arxiv on: 28 Jun 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper explores the use of various machine learning models to predict patient readmission rates due to non-adherence to scheduled follow-ups in patients with diabetes mellitus. The study utilizes the Diabetes 130-US Hospitals dataset, preprocessing the data before training traditional models such as XGBoost, LightGBM, CatBoost, Decision Tree, and Random Forest, as well as an in-house LSTM neural network for comparison. Performance evaluation was based on accuracy, precision, recall, and F1-score, with LightGBM emerging as the top-performing traditional model. The LSTM model showed high training accuracy but suffered from overfitting. SHAP values were used to improve model interpretability, identifying key factors such as number of lab procedures and discharge disposition as critical in predicting readmissions. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The study looks at how different machine learning models can help predict when patients with diabetes might not show up for follow-up appointments. Researchers used a big dataset about people with diabetes to train these models. They tested models like XGBoost, LightGBM, and Random Forest, as well as their own custom-made LSTM model. The best-performing model was LightGBM. While the LSTM model did well during training, it got stuck when trying to generalize to new data. By looking at how each model made its predictions, researchers found that things like lab test results and where patients went after leaving the hospital were important factors in predicting non-adherence. |
Keywords
» Artificial intelligence » Decision tree » F1 score » Lstm » Machine learning » Neural network » Overfitting » Precision » Random forest » Recall » Xgboost
