Black Magic in Deep Learning: How Human Skill Impacts Network Training
This addresses the problem of human skill variability impacting reproducibility and SOTA comparisons in deep learning, though it is incremental as an initial study.
The study investigated how a user's prior experience in deep learning affects hyperparameter optimization accuracy, finding a strong positive correlation where experienced participants achieved better performance with fewer resources, while novices followed random strategies.
How does a user's prior experience with deep learning impact accuracy? We present an initial study based on 31 participants with different levels of experience. Their task is to perform hyperparameter optimization for a given deep learning architecture. The results show a strong positive correlation between the participant's experience and the final performance. They additionally indicate that an experienced participant finds better solutions using fewer resources on average. The data suggests furthermore that participants with no prior experience follow random strategies in their pursuit of optimal hyperparameters. Our study investigates the subjective human factor in comparisons of state of the art results and scientific reproducibility in deep learning.