Hide-and-Seek: A Template for Explainable AI
This addresses the problem of neural network opacity for industry adoption, though it appears incremental as it builds on existing interpretability ideas.
The study tackles the lack of transparency in neural networks by proposing the Hide-and-Seek framework for training interpretable neural networks, achieving high interpretability without sacrificing predictive power.
Lack of transparency has been the Achilles heal of Neural Networks and their wider adoption in industry. Despite significant interest this shortcoming has not been adequately addressed. This study proposes a novel framework called Hide-and-Seek (HnS) for training Interpretable Neural Networks and establishes a theoretical foundation for exploring and comparing similar ideas. Extensive experimentation indicates that a high degree of interpretability can be imputed into Neural Networks, without sacrificing their predictive power.