Using machine learning to identify top predictors for nurses’ willingness to report medication errors

Array

Amany Farag¹  Amaury Lendasse²  Kaj-Mikael Björk³  Renjie Hu⁴ 

[1] Corresponding author.;Arcada University of Applied Sciences, Helsinki, Finland;College of Nursing, The University of Iowa, Iowa City, USA;Department of Industrial and Systems Engineering, The University of Iowa, Iowa City, USA;

关键词: Variable selection;    Data visualization;    Dimensionality reduction;    Extreme learning machines;    Self-organizing maps;    Medication error reporting;   

DOI  :  

来源: DOAJ

【 摘 要 】

This paper presents a novel methodology to analyze nurses’ willingness to report medication errors. Parallel Extreme Learning Machines were applied to identify the top interpersonal and organizational predictors and Self-Organizing Maps to create comprehensive visualization. The results of the data analysis were targeted to improve the likelihood of nurses reporting of medication errors. ELMs are accurate by extremely fast prediction models. Self-Organizing Maps enable us to perform non-linear dimensionality reduction to get an accurate visualization of the selected variables. Combining both techniques reduces the curse of dimensionality and improves the interpretability of the visualization.

【 授权许可】

Unknown