Risk Prediction Method of Obstetric Nursing Based on Data Mining

Area of Science:

• Nursing
• Data Science
• Medical Informatics

Background:

• Obstetric nursing involves inherent complexities and risks impacting patient safety and hospital outcomes.
• Effective risk prediction and timely intervention are crucial for optimal patient recovery in obstetric care.

Purpose of the Study:

• To develop and evaluate a novel forecasting model for predicting risks in obstetric nursing.
• To enhance the accuracy and stability of risk prediction compared to single-model approaches.

Main Methods:

• Utilized data mining techniques, specifically combining Support Vector Machine (SVM) and XGBoost algorithms.
• Developed a hybrid SVM-XGBoost prediction model to leverage the strengths of both machine learning methods.

Main Results:

• The combined SVM-XGBoost model achieved a prediction accuracy of 100%.
• This represents a significant improvement over single SVM (approx. 78% accuracy) and XGBoost (approx. 75% accuracy) models.
• The hybrid model also demonstrated enhanced precision and recall rates.

Conclusions:

• The SVM-XGBoost combined prediction model is highly effective and suitable for predicting obstetric nursing risks.
• This approach offers a robust solution for improving patient safety and care quality in obstetrics.