Papers
Topics
Authors
Recent
Search
2000 character limit reached

Dual feature-based and example-based explanation methods

Published 29 Jan 2024 in cs.LG, cs.AI, and stat.ML | (2401.16294v1)

Abstract: A new approach to the local and global explanation is proposed. It is based on selecting a convex hull constructed for the finite number of points around an explained instance. The convex hull allows us to consider a dual representation of instances in the form of convex combinations of extreme points of a produced polytope. Instead of perturbing new instances in the Euclidean feature space, vectors of convex combination coefficients are uniformly generated from the unit simplex, and they form a new dual dataset. A dual linear surrogate model is trained on the dual dataset. The explanation feature importance values are computed by means of simple matrix calculations. The approach can be regarded as a modification of the well-known model LIME. The dual representation inherently allows us to get the example-based explanation. The neural additive model is also considered as a tool for implementing the example-based explanation approach. Many numerical experiments with real datasets are performed for studying the approach. The code of proposed algorithms is available.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (77)
  1. One explanation does not fit all: A toolkit and taxonomy of AI explainability techniques. arXiv:1909.03012, Sep 2019.
  2. V. Belle and I. Papantonis. Principles and practice of explainable machine learning. Frontiers in Big Data, page 39, 2021.
  3. A survey of methods for explaining black box models. ACM computing surveys, 51(5):93, 2019.
  4. Explaining the black-box model: A survey of local interpretation methods for deep neural networks. Neurocomputing, 419:168–182, 2021.
  5. C. Molnar. Interpretable Machine Learning: A Guide for Making Black Box Models Explainable. Published online, https://christophm.github.io/interpretable-ml-book/, 2019.
  6. Interpretable machine learning: definitions, methods, and applications. arXiv:1901.04592, Jan 2019.
  7. Explainable deep learning: A field guide for the uninitiated. Journal of Artificial Intelligence Research, 73:329–396, 2022.
  8. Explainability of vision-based autonomous driving systems: Review and challenges. arXiv:2101.05307, January 2021.
  9. A survey on neural network interpretability. IEEE Transactions on Emerging Topics in Computational Intelligence, 5(5):726–742, 2021.
  10. “Why should I trust You?” Explaining the predictions of any classifier. In Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining, pages 1135–1144, 2016.
  11. Neural additive models: Interpretable machine learning with neural nets. Advances in neural information processing systems, 34:4699–4711, 2021.
  12. T. Hastie and R. Tibshirani. Generalized additive models, volume 43. CRC press, 1990.
  13. A unified approach to interpreting model predictions. In Advances in Neural Information Processing Systems, pages 4765–4774, 2017.
  14. E. Strumbelj and I. Kononenko. An efficient explanation of individual classifications using game theory. Journal of Machine Learning Research, 11:1–18, 2010.
  15. L.S. Shapley. A value for n-person games. In Contributions to the Theory of Games, volume II of Annals of Mathematics Studies 28, pages 307–317. Princeton University Press, Princeton, 1953.
  16. E. Strumbelj and I. Kononenko. A general method for visualizing and explaining black-box regression models. In Adaptive and Natural Computing Algorithms. ICANNGA 2011, volume 6594 of Lecture Notes in Computer Science, pages 21–30, Berlin, 2011. Springer.
  17. E. Strumbelj and I. Kononenko. Explaining prediction models and individual predictions with feature contributions. Knowledge and Information Systems, 41:647–665, 2014.
  18. Ensembles of random SHAPs. Algorithms, 15(11, 431):1–27, 2022.
  19. R. Yousefzadeh. Deep learning generalization and the convex hull of training sets. In NeurIPS 2020 Workshop: Deep Learning through Information Geometry, pages 1–10, 2020.
  20. Simulation and the Monte Carlo method, 2nd Edition. Wiley, New Jersey, 2008.
  21. Sampling uniformly from the unit simplex. Technical Report 29, Johns Hopkins University, 2004.
  22. Visualizing the effects of predictor variables in black box supervised learning models. Journal of the Royal Statistical Society: Series B (Statistical Methodology), 82(4):1059–1086, 2020.
  23. Alime: Autoencoder based approach for local interpretability. In Intelligent Data Engineering and Automated Learning–IDEAL 2019: 20th International Conference, Manchester, UK, November 14–16, 2019, Proceedings, Part I 20, pages 454–463. Springer, 2019.
  24. Anchors: High-precision model-agnostic explanations. In AAAI Conference on Artificial Intelligence, pages 1527–1535, 2018.
  25. Enriching visual with verbal explanations for relational concepts: Combining LIME with Aleph. In Machine Learning and Knowledge Discovery in Databases: International Workshops of ECML PKDD 2019, pages 180–192. Springer, 2020.
  26. SurvLIME: A method for explaining machine learning survival models. Knowledge-Based Systems, 203:106164, 2020.
  27. D. Garreau and U. von Luxburg. Explaining the explainer: A first theoretical analysis of LIME. In International conference on artificial intelligence and statistics, pages 1287–1296. PMLR, 2020.
  28. D. Garreau and U. von Luxburg. Looking deeper into tabular LIME. arXiv:2008.11092, August 2020.
  29. GraphLIME: Local interpretable model explanations for graph neural networks. IEEE Transactions on Knowledge and Data Engineering, 35(7):6968–6972, 2022.
  30. Gami-net: An explainable neural networkbased on generalized additive models with structured interactions. Pattern Recognition, 120:108192, 2021.
  31. Adaptive explainable neural networks (AxNNs). arXiv:2004.02353v2, April 2020.
  32. InterpretML: A unified framework for machine learning interpretability. arXiv:1909.09223, September 2019.
  33. Interpretable machine learning with an ensemble of gradient boosting machines. Knowledge-Based Systems, 222(106993):1–16, 2021.
  34. FastSHAP: Real-time shapley value estimation. In The Tenth International Conference on Learning Representations, ICLR 2022, pages 1–23, 2022.
  35. On locality of local explanation models. In Advances in neural information processing systems, volume 34, pages 18395–18407, 2021.
  36. SHAFF: Fast and consistent SHApley eFfect estimates via random Forests. In International Conference on Artificial Intelligence and Statistics, pages 5563–5582. PMLR, 2022.
  37. TimeSHAP: Explaining recurrent models through sequence perturbations. In Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining, pages 2565–2573, 2021.
  38. X-SHAP: towards multiplicative explainability of machine learning. arXiv:2006.04574, June 2020.
  39. Shapley explanation networks. arXiv:2104.02297, Apr 2021.
  40. R. Fong and A. Vedaldi. Explanations for attributing deep neural network predictions. In Explainable AI, volume 11700 of LNCS, pages 149–167. Springer, Cham, 2019.
  41. R.C. Fong and A. Vedaldi. Interpretable explanations of black boxes by meaningful perturbation. In Proceedings of the IEEE International Conference on Computer Vision, pages 3429–3437. IEEE, 2017.
  42. RISE: Randomized input sampling for explanation of black-box models. arXiv:1806.07421, June 2018.
  43. Evaluating explainers via perturbation. arXiv:1906.02032v1, Jun 2019.
  44. Techniques for interpretable machine learning. Communications of the ACM, 63(1):68–77, 2019.
  45. LEAFAGE: Example-based and feature importance-based explanations for black-box ML models. In 2019 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE), pages 1–7. IEEE, 2019.
  46. The effects of example-based explanations in a machine learning interface. In Proceedings of the 24th International Conference on Intelligent User Interfaces, pages 258–262, 2019.
  47. Toward scalable and unified example-based explanation and outlier detection. IEEE Trans Image Process, 31:525–540, 2022.
  48. Explaining latent representations with a corpus of examples. In Advances in Neural Information Processing Systems, volume 34, pages 12154–12166, 2021.
  49. Interactive label cleaning with example-based explanations. In Advances in Neural Information Processing Systems, volume 34, pages 12966–12977, 2021.
  50. Axiomatic attribution for deep networks. In 34th International Conference on Machine Learning, ICML, volume 7, pages 5109–5118, 2017.
  51. Explanations based on the missing: Towards contrastive explanations with pertinent negatives. arXiv:1802.07623v2, Oct 2018.
  52. A. Adadi and M. Berrada. Peeking inside the black-box: A survey on explainable artificial intelligence (XAI). IEEE Access, 6:52138–52160, 2018.
  53. Explainable artificial intelligence (XAI): Concepts, taxonomies, opportunities and challenges toward responsible AI. Information Fusion, 58:82–115, 2020.
  54. Benchmarking and survey of explanation methods for black box models. Data Mining and Knowledge Discovery, 2023.
  55. N. Burkart and M.F. Huber. A survey on the explainability of supervised machine learning. Journal of Artificial Intelligence Research, 70:245–317, 2021.
  56. Machine learning interpretability: A survey on methods and metrics. Electronics, 8(832):1–34, 2019.
  57. A systematic review of explainable artificial intelligence in terms of different application domains and tasks. Applied Sciences, 12(3:1353):1–38, 2022.
  58. Interpretable deep learning: Interpretations, interpretability, trustworthiness, and beyond. Knowledge and Information Systems, 64(12):3197–3234, 2022.
  59. C. Rudin. Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead. Nature Machine Intelligence, 1:206–215, 2019.
  60. Interpretable machine learning: Fundamental principles and 10 grand challenges. arXiv:2103.11251, March 2021.
  61. Large data sets classification using convex-concave hull and support vector machine. Soft Computing, 17:793–804, 2013.
  62. Extreme vector machine for training on large data. International Journal of Machine Learning and Cybernetics, 11:33–53, 2020.
  63. Machine learning algorithm based on convex hull analysis. Procedia Computer Science, 186:381–386, 2021.
  64. Nearest convex hull classification based on linear programming. Pattern Recognition and Image Analysis, 31:205–211, 2021.
  65. Online support vector machine based on convex hull vertices selection. IEEE Transactions on Neural Networks and Learning Systems, 24(4):593–609, 2013.
  66. Learning in high dimension always amounts to extrapolation. arXiv:2110.09485v2, Oct 2021.
  67. A convex hull-based data selection method for data driven models. Applied Soft Computing, 47:515–533, 2016.
  68. Duality and geometry in svm classifiers. In ICML’00: Proceedings of the Seventeenth International Conference on Machine Learning, pages 57–64, 2000.
  69. T. Zhang. On the dual formulation of regularized linear systems with convex risks. Machine Learning, 46(1):91–129, 2002.
  70. T. Ergen and M. Pilanci. Convex duality of deep neural networks. In Proceedings of the 37 th International Conference on Machine Learning, volume PMLR 108, 2020.
  71. T. Ergen and M. Pilanci. Convex geometry and duality of over-parameterized neural networks. Journal of Machine Learning Research, 22(212):1–63, 2021.
  72. F. Farnia and D. Tse. A convex duality framework for gans. In 32nd Conference on Neural Information Processing Systems (NeurIPS 2018), pages 1–11, 2018.
  73. High-dimensional similarity learning via dual-sparse random projection. In Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence (IJCAI-18), pages 3005–3011, 2018.
  74. R.T. Rockafellar. Convex Analysis, volume 28 of Princeton Mathematical Series. Princeton University Press, Princeton, N.J., 1970.
  75. Mixup: Beyond empirical risk minimization. In Proceedings of ICLR, pages 1–13, 2018.
  76. J.H. Friedman. Greedy function approximation: A gradient boosting machine. Annals of Statistics, 29:1189–1232, 2001.
  77. A simple and effective model-based variable importance measure. arXiv:1805.04755, May 2018.

Summary

No one has generated a summary of this paper yet.

Sign Up to Summarize

Paper to Video (Beta)

No one has generated a video about this paper yet.

Sign Up to Generate All Videos Subscribe on YouTube

Whiteboard

No one has generated a whiteboard explanation for this paper yet.

Sign Up to Generate

Open Problems

We haven't generated a list of open problems mentioned in this paper yet.

Generate Now

Continue Learning

We haven't generated follow-up questions for this paper yet.

Generate Now

Collections

Sign up for free to add this paper to one or more collections.

Sign Up

Tweets

Sign up for free to view the 2 tweets with 0 likes about this paper.

Sign Up for Free