Balancing Explainability-Accuracy of Complex Models
This addresses the problem of explainable AI for complex models in domains like healthcare and autonomous driving, but it appears incremental as it builds on existing XAI research.
The paper tackles the explainability-accuracy tradeoff in complex models like neural networks by introducing a new approach based on co-relation impact, which enhances explainability while maintaining high accuracy and provides a computational complexity bound for dependent features.
Explainability of AI models is an important topic that can have a significant impact in all domains and applications from autonomous driving to healthcare. The existing approaches to explainable AI (XAI) are mainly limited to simple machine learning algorithms, and the research regarding the explainability-accuracy tradeoff is still in its infancy especially when we are concerned about complex machine learning techniques like neural networks and deep learning (DL). In this work, we introduce a new approach for complex models based on the co-relation impact which enhances the explainability considerably while also ensuring the accuracy at a high level. We propose approaches for both scenarios of independent features and dependent features. In addition, we study the uncertainty associated with features and output. Furthermore, we provide an upper bound of the computation complexity of our proposed approach for the dependent features. The complexity bound depends on the order of logarithmic of the number of observations which provides a reliable result considering the higher dimension of dependent feature space with a smaller number of observations.