Enhancing Performance of Explainable AI Models with Constrained Concept Refinement
This work addresses the challenge of balancing accuracy and interpretability in machine learning, particularly for emerging interpretable-by-design methods, which is significant for researchers and practitioners seeking trustworthy and reliable AI models.
The authors tackled the trade-off between accuracy and interpretability in machine learning, achieving zero loss while enhancing model interpretability, and improving prediction accuracy with lower computational cost. Their approach outperformed existing explainable methods across various large-scale benchmarks.
The trade-off between accuracy and interpretability has long been a challenge in machine learning (ML). This tension is particularly significant for emerging interpretable-by-design methods, which aim to redesign ML algorithms for trustworthy interpretability but often sacrifice accuracy in the process. In this paper, we address this gap by investigating the impact of deviations in concept representations-an essential component of interpretable models-on prediction performance and propose a novel framework to mitigate these effects. The framework builds on the principle of optimizing concept embeddings under constraints that preserve interpretability. Using a generative model as a test-bed, we rigorously prove that our algorithm achieves zero loss while progressively enhancing the interpretability of the resulting model. Additionally, we evaluate the practical performance of our proposed framework in generating explainable predictions for image classification tasks across various benchmarks. Compared to existing explainable methods, our approach not only improves prediction accuracy while preserving model interpretability across various large-scale benchmarks but also achieves this with significantly lower computational cost.