PulseTrendingJournal ClubResearchersJournalsExplore
Instagram
HomeTrendingJournal ClubExplore
Synapse
⌘+K
Synapse
April 22, 2026AIOpen Access

Comparative Forecasting and Misclassification Analysis Using Health Survey Data

View Full Paper
Ask AI
Bookmark
Share

Authors

ETErmioni TrakaGPGeorge PapageorgiouGMGeorgios Mantzavinis

Discussion

Loading...

Member takes

Overview

Comparative analysis reveals improved mortality prediction using machine learning in health surveys, indicating novel approaches.

Key Points

  • This study aims to enhance mortality prediction accuracy by exploring machine learning techniques and addressing misclassification issues.
  • Utilized IPUMS-NHIS datasets from 2010 and 2015 for analysis.
  • Employed SMOTE for class imbalance correction and hyperparameter tuning for model optimization.
  • Trained Logistic Regression, Random Forest, and XGBoost models to classify mortality.
  • XGBoost outperformed other models with a recall of 69% and AUC of 0.92.
  • Key predictors identified include age, employment status, and self-reported health.
  • Atypical profiles not captured by standard models were revealed through misclassification analysis.

Cite This Study

Traka et al. (2026) studied this question.

synapsesocial.com/papers/69e866c96e0dea528ddeb2d1https://doi.org/10.3390/ai7040148
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1Data balancing improves mortality prediction for emergency department patients2026
  2. 2Using tree-based ensemble methods to produce a population-based mortality risk score in Ontario, Canada2026
  3. 3Machine Learning Models for Predicting COVID-19 Mortality Using Epidemiological Features2025
  4. 4Optimizing Mortality Prediction for ICU Heart Failure Patients: Leveraging XGBoost and Advanced Machine Learning with the MIMIC-III Database2024 · 2 citations
  5. 5Optimizing stroke mortality prediction: a comparative study of sampling, cost-sensitive learning, and threshold adjustment2026