@article{ref1,
title="Comparison of artificial neural network and logistic regression models for prediction of psychological symptom six months after mild traumatic brain injury",
journal="Iranian journal of psychiatry and behavioral sciences",
year="2017",
author="Shafiei, Elham and Fakharian, Esmaeil and Omidi, Abdollah and Akbari, Hossein and Delpisheh, Ali and Nademi, Arash",
volume="11",
number="3",
pages="e5849-e5849",
abstract="BACKGROUND: Nowadays, outcome prediction models using logistic regression (LR) and artificial neural network (ANN) analysis have been developed in many areas of healthcare research.   Objectives: In this study, we have compared the performance of multivariable LR and ANN models, in prediction of psychological symptoms six months after mild traumatic brain injury.   Methods: In a prospective cohort study, information of 100 mild traumatic brain injury patients, during a six months period between 2014 and 2016 were included. Data were divided into two training (n = 50) and testing (n = 50) groups, randomly. 300 ANNs and LRs were studied in the first group and then the predicted values were compared in the second group using the two final models. The receiver operating characteristic (ROC) curve and accuracy rate were used to compare these models.   Results: The results showed that accuracy rate for the neural network model was 90.65%, while it was 75.96% for the LR model.   Conclusions: The ANN models appeared to be more powerful in predicting psychological symptoms versus the LR models.   Keywords: Artificial Neural Network; Logistic Regression; Mental Disorder; Mild Traumatic Brain Injury; Prediction; Principle Component Analysis; Psychological Symptom  Copyright © 2016, Mazandaran University of Medical Sciences. This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/) which permits copy and redistribute the material just in noncommercial usages, provided the original work is properly cited.<p /> <p>Language: en</p>",
language="en",
issn="1735-8639",
doi="10.17795/ijpbs-5849",
url="http://dx.doi.org/10.17795/ijpbs-5849"
}