Delivering World-Class Science
CHAI will address four fundamental research challenges:
Causality for ML | Enhancing robustness of predictive models against heterogeneity, distribution shifts, measurement error (noise) and fairness issues by exploiting causal principles.
Large scale causal discovery | Inferring invariances to domain shifts and interventions across large and diverse data regimes and missingness patterns (sparsity) to inform causal models.
Causal representation learning | Building deep representations from unstructured data (e.g. images, text) optimised for detecting effects of interventions and data and domain shifts.
Causal reasoning for decision support | Deriving optimal policies integrating actions, prediction under interventions and risk modelling accounting for data-constrained uncertainty.
Publications
2024
Interactive molecular causal networks of hypertension using a fast machine learning algorithm MRdualPC Jack Kelly, Xiaoguang Xu, James M. Eales, Bernard Keavney, Carlo Berzuini, Maciej Tomaszewski & Hui Guo
Bounding Causal Effects with Leaky Instruments David S. Watson, Jordan Penn, Lee M. Gunderson, Gecia Bravo-Hermsdorff, Afsaneh Mastouri, Ricardo Silva
https://arxiv.org/abs/2404.04446
Counterfactual contrastive learning: robust representations via causal image synthesis Melanie Roschewitz, Fabio De Sousa Ribeiro, Tian Xia, Galvin Khara, Ben Glocker
Mitigating attribute amplification in counterfactual image generation, Tian Xia, Mélanie Roschewitz, Fabio De Sousa Ribeiro, Charles Jones, Ben Glocker,
https://arxiv.org/abs/2403.09422
Fairtune: Optimizing parameter efficient fine tuning for fairness in medical image analysis. Dutt, R., Bohdal, O., Tsaftaris, S. A., & Hospedales, T. (2024). International Conference on Learning Representations. https://openreview.net/forum?id=ArpwmicoYW
Causal blind spots when using prediction models for treatment decisions. van Geloven, N., Keogh, R. H., van Amsterdam, W., Cinà, G., Krijthe, J. H., Peek, N., …, Sperrin, M., ... & Didelez, V. (2024). arXiv preprint arXiv:2402.17366. https://arxiv.org/abs/2402.17366
Pragmatic fairness: Developing policies with outcome disparity control. Gultchin, L., Guo, S., Malek, A., Chiappa, S., & Silva, R. (2024). Causal Learning and Reasoning. PMLR. https://arxiv.org/abs/2301.12278
Making predictions under interventions: a case study from the PREDICT-CVD cohort in New Zealand primary care. Lin, L., Poppe, K., Wood, A., Martin, G. P., Peek, N., & Sperrin, M. (2024). Frontiers in Epidemiology, 4, 1326306. https://doi.org/10.3389/fepid.2024.1326306
A causal perspective on dataset bias in machine learning for medical imaging. Jones, C., Castro, D. C., De Sousa Ribeiro, F., Oktay, O., McCradden, M., & Glocker, B. (2024). Nature Machine Intelligence. https://doi.org/10.1038/s42256-024-00797-8
2023
A Causal Ordering Prior for Unsupervised Representation Learning. Kori, A., Sanchez, P., Vilouras, K., Glocker, B., & Tsaftaris, S. A. (2023). arXiv preprint arXiv:2307.05704. https://arxiv.org/abs/2307.05704
Intervention generalization: A view from factor graph models. Bravo-Hermsdorff, G., Watson, D., Yu, J., Zeitler, J., & Silva, R. (2023). Advances in Neural Information Processing Systems, 36. https://arxiv.org/abs/2306.04027
Risk-based decision making: estimands for sequential prediction under interventions. Luijken, K., Morzywołek, P., van Amsterdam, W., Cinà, G., Hoogland, J., Keogh, R., …, Sperrin, M., ... & van Geloven, N. (2023). arXiv preprint arXiv:2311.17547. https://arxiv.org/abs/2311.17547
Stochastic causal programming for bounding treatment effects. Padh, K., Zeitler, J., Watson, D., Kusner, M., Silva, R., & Kilbertus, N. (2023). Causal Learning and Reasoning. PMLR. https://arxiv.org/abs/2202.10806
Challenges in Explaining Brain Tumor Detection. Legastelois, B., Rafferty, A., Brennan, P., Chockler, H., Rajan, A., & Belle, V. (2023). International Symposium on Trustworthy Autonomous Systems. https://doi.org/10.1145/3597512.3600208
High fidelity image counterfactuals with probabilistic causal models. Ribeiro, F. D. S., Xia, T., Monteiro, M., Pawlowski, N., & Glocker, B. (2023). International Conference on Machine Learning. https://openreview.net/forum?id=DA0PROpwan
Measuring axiomatic soundness of counterfactual image models. Monteiro, M., Ribeiro, F. D. S., Pawlowski, N., Castro, D. C., & Glocker, B. (2023). International Conference on Learning Representations. https://openreview.net/forum?id=lZOUQQvwI3q
2022
Diffusion models for causal discovery via topological ordering. Sanchez, P., Liu, X., O'Neil, A. Q., & Tsaftaris, S. A. (2022). International Conference on Learning Representations. https://openreview.net/forum?id=Idusfje4-Wq
Targeted validation: validating clinical prediction models in their intended population and setting. Sperrin, M., Riley, R. D., Collins, G. S., & Martin, G. P. (2022). Diagnostic and Prognostic Research, 6(24). https://doi.org/10.1186/s41512-022-00136-8
Causal inference with treatment measurement error: a nonparametric instrumental variable approach. Zhu, Y., Gultchin, L., Gretton, A., Kusner, M. J., & Silva, R. (2022). Uncertainty in Artificial Intelligence. PMLR. https://arxiv.org/abs/2206.09186
Diffusion causal models for counterfactual estimation. Sanchez, P., & Tsaftaris, S. A. (2022). Causal Learning and Reasoning. PMLR. https://arxiv.org/abs/2202.10166
Causal discovery under a confounder blanket. Watson, D. S., & Silva, R. (2022). Uncertainty in Artificial Intelligence. PMLR. https://arxiv.org/abs/2205.05715
Causal machine learning for healthcare and precision medicine. Sanchez, P., Voisey, J. P., Xia, T., Watson, H. I., O’Neil, A. Q., & Tsaftaris, S. A. (2022). Royal Society Open Science, 9(8), 220638. https://doi.org/10.1098/rsos.220638
Looking before we leap: Expanding ethical review processes for AI and data science research. Petermann, M., Tempini, N., Kherroubi Garcia, I., Whitaker, K., & Strait, A. (2022). Ada Lovelace Institute, London. ISBN 978-1-7397950-6-1. https://www.adalovelaceinstitute.org/report/looking-before-we-leap/
Data Circulation in Health Landscapes. Tempini, N., Maturo, A., & Tola, E. (2022). Tecnoscienza–Italian Journal of Science & Technology Studies, 13(1). https://doi.org/10.6092/issn.2038-3460/17567
Potential sources of dataset bias complicate investigation of underdiagnosis by machine learning algorithms. Bernhardt, M., Jones, C., & Glocker, B. (2022). Nature Medicine, 28(6). https://doi.org/10.1038/s41591-022-01846-8
Quantifying harm. Beckers, S., Chockler, H., & Halpern, J. Y. (2022). International Joint Conference on Artificial Intelligence. https://arxiv.org/abs/2209.15111
On testing for discrimination using causal models. Chockler, H., & Halpern, J. Y. (2022). AAAI Conference on Artificial Intelligence (Vol. 36, No. 5). https://doi.org/10.1609/aaai.v36i5.20494
A causal analysis of harm. Beckers, S., Chockler, H., & Halpern, J. (2022). Advances in Neural Information Processing Systems, 35. https://arxiv.org/abs/2210.05327
2021
Explanations for occluded images. Chockler, H., Kroening, D., & Sun, Y. (2021). IEEE/CVF International Conference on Computer Vision. https://arxiv.org/abs/2103.03622
Ranking policy decisions. Pouget, H., Chockler, H., Sun, Y., & Kroening, D. (2021). Advances in Neural Information Processing Systems, 34. https://arxiv.org/abs/2008.13607
A scoping review of causal methods enabling predictions under hypothetical interventions. Lin, L., Sperrin, M., Jenkins, D. A., Martin, G. P., & Peek, N. (2021). Diagnostic and Prognostic Research, 5(3). https://doi.org/10.1186/s41512-021-00092-9
Invited commentary: treatment drop-in—making the case for causal prediction. Sperrin, M., Diaz-Ordaz, K., & Pajouheshnia, R. (2021). American Journal of Epidemiology, 190(10). https://doi.org/10.1093/aje/kwab030
Data curation-research: practices of data standardization and exploration in a precision medicine database. Tempini, N. (2021). New Genetics and Society, 40(1). https://doi.org/10.1080/14636778.2020.1853513
Actionable data for precision oncology: Framing trustworthy evidence for exploratory research and clinical diagnostics. Tempini, N., & Leonelli, S. (2021). Social Science & Medicine, 272, 113760. https://doi.org/10.1016/j.socscimed.2021.113760
Causality in digital medicine. Glocker, B., Musolesi, M., Richens, J., & Uhler, C. (2021). Nature Communications, 12(1). https://doi.org/10.1038/s41467-021-25743-9
2020
Data journeys in the sciences. Leonelli, S., & Tempini, N. (2020). Springer Nature. https://doi.org/10.1007/978-3-030-37177-7
Deep structural causal models for tractable counterfactual inference. Pawlowski, N., Castro, D. C., & Glocker, B. (2020). Advances in Neural Information Processing Systems, 33. https://arxiv.org/abs/2006.06485
Causality matters in medical imaging. Castro, D. C., Walker, I., & Glocker, B. (2020). Nature Communications, 11(1). https://doi.org/10.1038/s41467-020-17478-w
Combining experts' causal judgments. Alrajeh, D., Chockler, H., & Halpern, J. Y. (2020). Artificial Intelligence, 288, 103355. https://doi.org/10.1016/j.artint.2020.103355
Learning the language of software errors. Chockler, H., Kesseli, P., Kroening, D., & Strichman, O. (2020). Journal of Artificial Intelligence Research, 67. https://doi.org/10.1613/jair.1.11798
Explaining image classifiers using statistical fault localization. Sun, Y., Chockler, H., Huang, X., & Kroening, D. (2020, August). European Conference on Computer Vision. https://doi.org/10.1007/978-3-030-58604-1_24