Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
103 tokens/sec
GPT-4o
11 tokens/sec
Gemini 2.5 Pro Pro
50 tokens/sec
o3 Pro
5 tokens/sec
GPT-4.1 Pro
3 tokens/sec
DeepSeek R1 via Azure Pro
33 tokens/sec
2000 character limit reached

Incorporating Expert Rules into Neural Networks in the Framework of Concept-Based Learning (2402.14726v1)

Published 22 Feb 2024 in cs.LG, cs.AI, and stat.ML

Abstract: A problem of incorporating the expert rules into machine learning models for extending the concept-based learning is formulated in the paper. It is proposed how to combine logical rules and neural networks predicting the concept probabilities. The first idea behind the combination is to form constraints for a joint probability distribution over all combinations of concept values to satisfy the expert rules. The second idea is to represent a feasible set of probability distributions in the form of a convex polytope and to use its vertices or faces. We provide several approaches for solving the stated problem and for training neural networks which guarantee that the output probabilities of concepts would not violate the expert rules. The solution of the problem can be viewed as a way for combining the inductive and deductive learning. Expert rules are used in a broader sense when any logical function that connects concepts and class labels or just concepts with each other can be regarded as a rule. This feature significantly expands the class of the proposed results. Numerical examples illustrate the approaches. The code of proposed algorithms is publicly available.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (43)
  1. Learning interpretable concept-based models with human feedback. arXiv:2012.02898, Dec 2020.
  2. Learning bottleneck concepts in image classification. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 10962–10971, 2023.
  3. Interpretability beyond feature attribution: Quantitative testing with concept activation vectors (tcav). In International conference on machine learning, pages 2668–2677. PMLR, 2018.
  4. On completeness-aware concept-based explanations in deep neural networks. In Advances in neural information processing systems, volume 33, pages 20554–20565, 2020.
  5. Concept bottleneck models. In International conference on machine learning, pages 5338–5348. PMLR, 2020.
  6. Concept embedding models: Beyond the accuracy-explainability trade-off. In Advances in Neural Information Processing Systems, volume 35, pages 21400–21413, 2022.
  7. The caltech-ucsd birds-200-2011 dataset. Technical report, Technical report, California Institute of Technology, 2011.
  8. Concept-based explanations for tabular data. arXiv:2209.05690, Sep 2022.
  9. Visual concept learning: Combining machine vision and bayesian generalization on concept hierarchies. In Advances in Neural Information Processing Systems, volume 26, pages 1–9, 2013.
  10. Understanding the (extra-) ordinary: Validating deep model decisions with prototypical concept-based explanations. arXiv:2311.16681, Nov 2023.
  11. Concept correlation and its effects on concept-based models. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, pages 4780–4788, 2023.
  12. Human uncertainty in concept-based ai systems. In Proceedings of the 2023 AAAI/ACM Conference on AI, Ethics, and Society, pages 869–889, 2023.
  13. Statistically significant concept-based explanation of image classifiers via model knockoffs. arXiv:2305.18362, May 2023.
  14. Interpretable and intervenable ultrasonography-based machine learning models for pediatric appendicitis. Medical Image Analysis, 91:103042, 2024.
  15. The natural language explanation algorithms for the lung cancer computer-aided diagnosis system. Artificial Intelligence in Medicine, 108(Article 101952):1–10, 2020.
  16. Coherent concept-based explanations in medical image and its application to skin lesion diagnosis. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 3798–3807, 2023.
  17. Towards concept-based interpretability of skin lesion diagnosis using vision-language models. arXiv:2311.14339, Nov 2023.
  18. Robust and interpretable medical image classifiers via concept bottleneck models. arXiv:2310.03182, Oct 2023.
  19. Example or prototype? learning concept-based explanations in time-series. In Asian Conference on Machine Learning, pages 816–831. PMLR, 2023.
  20. Interpretable time-series classification on few-shot samples. In 2020 International Joint Conference on Neural Networks (IJCNN), pages 1–8. IEEE, 2020.
  21. Concept-based explanations for out-of-distribution detectors. In International Conference on Machine Learning, pages 5817–5837. PMLR, 2023.
  22. Transparent anomaly detection via concept-based explanations. arXiv:2310.10702, Oct 2023.
  23. Promises and pitfalls of black-box concept learning models. arXiv:2106.13314, Jun 2021.
  24. From neural activations to concepts: A survey on explaining concepts in neural networks. arXiv:2310.11884, Oct 2023.
  25. Post-hoc concept bottleneck models. arXiv:2205.15480, May 2022.
  26. I. Sheth and S.E. Kahou. Auxiliary losses for learning generalizable concept-based models. arXiv:2311.11108, Nov 2023.
  27. Probabilistic concept bottleneck models. arXiv:2306.01574, Jun 2023.
  28. Concept bottleneck generative models. In Proceedings of ICML 2023. Workshop on Deployment Challenges for Generative AI, https://openreview.net/group?id=ICML.cc/2023/Workshop, pages 1–10, 2023.
  29. Do concept bottleneck models obey locality? arXiv:2401.01259, Jan 2024.
  30. Do concept bottleneck models learn as intended? arXiv:2105.04289, May 2021.
  31. Beyond concept bottleneck models: How to make black boxes intervenable? arXiv:2401.13544, Jan 2024.
  32. Ceir: Concept-based explainable image representation learning. arXiv:2312.10747, Dec 2023.
  33. Concept bottleneck model with additional unsupervised concepts. IEEE Access, 10:41758–41765, 2022.
  34. Energy-based concept bottleneck models: unifying prediction, concept intervention, and conditional interpretations. arXiv:2401.14142, Jan 2024.
  35. Interactive concept bottleneck models. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 37, pages 5948–5955, 2023.
  36. Hierarchical concept bottleneck models for vision and their application to explainable fine classification and tracking. Engineering Applications of Artificial Intelligence, 118:105674, 2023.
  37. Addressing leakage in concept bottleneck models. In Advances in Neural Information Processing Systems, volume 35, pages 23386–23397, 2022.
  38. Eliminating information leakage in hard concept bottleneck models with supervised, hierarchical concept learning. arXiv:2402.05945, Feb 2024.
  39. Glancenets: Interpretable, leak-proof concept-based models. In Advances in Neural Information Processing Systems, volume 35, pages 21212–21227, 2022.
  40. Learning transferable visual models from natural language supervision. In International conference on machine learning, pages 8748–8763. PMLR, 2021.
  41. CLIP-QDA: An explainable concept bottleneck model. arXiv:2312.00110, Dec 2023.
  42. Komei Fukuda. Exact algorithms and software in optimization and polyhedral computation. In Proceedings of the Twenty-First International Symposium on Symbolic and Algebraic Computation, pages 333–334, 2008.
  43. A new computationally simple approach for implementing neural networks with output hard constraints. Doklady Mathematics, 2023.
Citations (2)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now
X Twitter Logo Streamline Icon: https://streamlinehq.com

Tweets