Predictive computing of human errors while training machine learning models

Due to the limited recognition of human beings, the training of machine learning models may cause uncertainties. Looking for the regularity of uncertain concepts is truly an issue that needs to be addressed. This study aims to investigate the human error modes and probabilities through training the machine learning model, especially for image pattern recognition under human-computer computing scenarios. The results showed that machine learning for image pattern recognition lacks reliability while training the models, and the professionalism of operators and the number of iterations are important to guarantee the system reliability.