ABOUT ML Reference Document

Last Updated

3.4.2.3 Evaluation

3.4.2.3 Evaluation

After a model is trained, it needs to be tested. The goal of documentation in this section is to help users of the model to understand how the model was checked for accuracy, bias, and other important properties.

The specific metrics that the model will be tested on depend on the particular use case, so it is helpful for this documentation to include examples of which metrics apply for which use case. Additionally, there will be contexts and use cases that are inappropriate for each model, so these limitations should be noted to avoid invalidating any fairness assertions.Selbst, Andrew D. and Boyd, Danah and Friedler, Sorelle and Venkatasubramanian, Suresh and Vertesi, Janet (2018). “Fairness and Abstraction in Sociotechnical Systems”, ACM Conference on Fairness, Accountability, and Transparency (FAT*). https://ssrn.com/abstract=3265913 Testing models should also take into account the broader system the model is embedded into, so it is valuable to test the downstream effects of the model. These anticipated downstream effects should be documented as a baseline to evaluate against on an ongoing basis.

Breaking down accuracy and other metrics by specific subpopulations can be useful for detecting possible biases and ethical issues. For example, if a model has much lower predictive accuracy or much higher false positive rates for one subpopulation, it might be problematic to use the model to make decisions for those subpopulations. The documentation should thus reflect those limitations clearly. Such disclosures would help ensure that the model is not later used to make unfair decisions for those subpopulations and could also give the developer some liability cushion against third-party misuse since they have clearly stated the limitations of their particular model. Part of the goal of this section is to convey to the user that whether AI works is not binary and instead is a nuanced question depending on the use case and metrics of success.

The documentation should also discuss how the model developers checked for overfitting. For example, how did they construct the test set? Did they draw new test sets each time they trained the model for cross-validation purposes? It is important for the documentation to be very specific and to explore potential shortcomings of the model and the test set. While the test set is meant to be a reflection of reality, data is never perfect. Failing to account for known imperfections creates a risk that this documentation could lead to overconfidence and misuse of the model.

Pros/Cons

Disclosing the details of evaluation puts the system at risk of being more easily manipulated by malicious actors, presenting a security risk. Indeed, the issue of Goodhart’s Law, as discussed previously, can be a concern if individuals deliberately change their behavior to try to change the model’s outcomes. That said, some activist groups engage in hacking to expose problematic aspects of an ML system with the goal of protecting vulnerable groups, so the increased potential for hacking can either be a positive or negative characteristic depending on one’s perspective. Explorations related to the following research questions could uncover insights into barriers to implementation along with mitigation strategies to overcome those barriers.

Sample Documentation Questions

  • Which datasets was the service tested on? (Arnold et al. 2018)
  • Describe the testing methodology. (Arnold et al. 2018)
  • Describe the test results. (Arnold et al. 2018)
  • Are you aware of possible examples of bias, ethical issues, or other safety risks as a result of using the service? (Arnold et al. 2018)
  • Are the service outputs explainable and/or interpretable? (Arnold et al. 2018)
  • Metrics. Metrics should be chosen to reflect potential real world impacts of the model. (Mitchell et al. 2018)
    • Model performance measures
    • Decision thresholds
    • Variation approaches
  • Evaluation data. Details on the dataset(s) used for the quantitative analyses in the card. (Mitchell et al. 2018)
    • Datasets
    • Motivation
    • Preprocessing

Promising Interventions

Those attempting documentation practices within any phase of the machine learning lifecycle can consider lessons learned from flawed ML systems currently in use by paying particular attention to:

  1. Instantiating more rigorous tests that include comparing system decisions with those a human would arrive at
  2. Putting in place a mitigation strategy for risks in the realm of possibility for homegrown systems

3.4.2.4 Model Integration

3.4.2.4 Model Integration

Model Integration

Integrating a model is the act of putting into production (deployment) and instantiating the monitoring process.

In a 2019 Towards Data Science blog post, Opeyemi notes that the deployment of an ML model “simply means the integration of the model into an existing production environment which can take in an input and return an output that can be used in making practical business decisions.”

Even if each portion of the model is thoroughly tested, validated, and documented, there are additional documentation and evaluation needs that arise when connecting a model into a broader ML system. Validating how the models interoperate is important because different models might not work well together. It is important to test for how errors from all the models interact to find the corner cases that could pose problems in production. Latency changes when connecting models into a system, and that changes usability and reliability for users as well. The logistics of how the model is run (e.g., on the cloud vs. on local machines) is another important factor to document because it changes how to handle client data in the pipeline, and clients in sensitive industries especially will be interested in the details of how that is handled. The system-level documentation should also include information about system logs, pre- and post-processing steps, and a summary of all the surrounding product and software design choices in the system which are not ML/AI related, but do impact how the ML/AI models operate in the system. Finally, there should be careful commentary on the system’s vulnerability to adversarial inputs and whether there are system-level mitigations that have been put into place. This information is obviously sensitive and should be carefully considered before release to minimize the risk of malicious use of this information.

Pros/Cons

System level information can be especially useful for consumers, helping them determine how the system can be used in their context. It can also help regulators examine otherwise black box systems for compliance on privacy and data storage regulations. Of course, this can also be seen as a liability for a company, but on the other hand, it can equally be a force for increasing compliance with regulations. An internal process that supports accurate documentation of the entire ML system will also mitigate the risk of inaccurate documentation creating legal liability. For example, if one team writes that the system does not log, but a different team upstream of them does, it would be inaccurate to publish that documentation. Fortunately, adhering to a rigorous internal auditing system before publishing documentation can catch this type of inaccuracy and will likely have the positive externalities of creating more cohesion between teams and preventing other issues that arise when functions become too siloed inside an organization. Explorations related to the following research questions could uncover insights into barriers to implementation along with mitigation strategies to overcome those barriers.

Sample Documentation Questions

  • What is the expected performance on unseen data or data with different distributions? (Arnold et al. 2018)
  • Was the service checked for robustness against adversarial attacks? (Arnold et al. 2018)
  • Quantitative analyses (Mitchell et al. 2018)
  • Unitary results
  • Intersectional results
  • Ethical considerations (Mitchell et al. 2018)

3.4.2.5 Maintenance

3.4.2.5 Maintenance

As data, techniques, and real-world needs change, most models must be updated to ensure continued usefulness. There are many parameters of a model update which would be useful to document, including how and why an update is triggered, whether old models or parameters can still be accessed, who owns maintenance and updating, and guidance on reasonable shelf life of the model, including performance expectations for old and new versions. The first question to answer is why the model is being updated: for underlying data changes, process changes, or to improve model performance? Most of the answers to the questions above flow from this first one.

Updates can be triggered automatically based on time or specific metrics, or be active decisions evaluated from time to time. It is important to be clear about this so that people using the model can plan their workflows accordingly.

When models are updated, it can be useful to maintain access to previous versions or parameters.

Along with information on who owns the maintenance and update process, it is useful to know what the fallback plan is in case of personnel turnover or organizational changes. Having this documentation well-known makes it easier to make sure that model maintenance does not fall through the cracks during transitions.

The shelf life of a model, similar to the update schedule itself, can be dependent on time only or a combination of other metrics or factors. For example, it could be that when the underlying data distribution looks X% different, then the model should be evaluated on Y criteria for whether it is still valid.

Pros/Cons

Completing all these steps of documenting an update process can be incredibly valuable both for people who will use the model and for the people building the model. For people using the model, they get more reliability and can better plan around the model based on update processes. For the people building the model, planning ahead for the update process can encourage an intentional approach and schedule up front. It also may encourage more precaution if the model developers have to document changes later, in the tradeoff between speed of progress and a precautionary principle.

There may be liability or timeline risks with thoroughly documenting the update process if the team misses deadlines or is locked into deadlines. One way to mitigate these risks would be to build in criteria/metric gates for updates rather than a hard timeline. For example, the update process might begin when X metric reaches Y level or one year from the publication date, whichever is earlier. Building in sufficient time to do the update is also important because some updates are simple but others can take a long time if they are larger updates. Explorations related to the following research questions could uncover insights into barriers to implementation along with mitigation strategies to overcome those barriers.

Sample Documentation Questions

  • Does the service implement and perform any bias detection and remediation? (Arnold et al. 2018)
  • When were the models last updated? (Arnold et al. 2018)

Promising Interventions

Those attempting documentation practices within any phase of the machine learning lifecycle can consider potential misuse by paying particular attention to:

  1. Determining the auditing method
  2. Archiving the old model (versioning for safety and recovery)

ABOUT ML Reference Document

Section 0: How to Use this Document

Recommended Reading Plan

Quick Guides

How We Define

Contact for Support

Section 1: Project Overview

1.1 Statement of Importance for ABOUT ML Project

1.1.0 Importance of Transparency: Why a Company Motivated by the Bottom Line Should Adopt ABOUT ML Recommendations

1.1.1 About This Document and Version Numbering

1.1.2 ABOUT ML Goals and Plan

1.1.3 ABOUT ML Project Process and Timeline Overview

1.1.4 Who Is This Project For?

1.1.4.1 Audiences for the ABOUT ML Resources

1.1.4.2 Stakeholders That Should Be Consulted While Putting Together ABOUT ML Resources

1.1.4.3 Audiences for ABOUT ML Documentation Artifacts

1.1.4.4 Whose Voices Are Currently Reflected in ABOUT ML?

1.1.4.5 Origin Story

Section 2: Literature Review (Current Recommendations on Documentation for Transparency in the ML Lifecycle)

2.1 Demand for Transparency and AI Ethics in ML Systems 

2.2 Documentation to Operationalize AI Ethics Goals

2.2.1 Documentation as a Process in the ML Lifecycle

2.2.2 Key Process Considerations for Documentation

2.3 Research Themes on Documentation for Transparency 

2.3.1 System Design and Set Up

2.3.2 System Development

2.3.3 System Deployment

Section 3: Preliminary Synthesized Documentation Suggestions

3.4.1 Suggested Documentation Sections for Datasets

3.4.1.1 Data Specification

3.4.1.1.1 Motivation

3.4.1.2 Data Curation 

3.4.1.2.1 Collection

3.4.1.2.2 Processing

3.4.1.2.3 Composition

3.4.1.2.4 Types and Sources of Judgement Calls

3.4.1.3 Data Integration

3.4.1.3.1 Use

3.4.1.3.2 Distribution

3.4.1.4 Maintenance

3.4.2 Suggested Documentation Sections for Models

3.4.2.1 Model Specifications

3.4.2.2 Model Training

3.4.2.3 Evaluation

3.4.2.4 Model Integration

3.4.2.5 Maintenance

Section 4: Current Challenges of Implementing Documentation

Section 5: Conclusions

Version 0

Version 1

Appendix A: Compiled List of Documentation Questions 

Fact Sheets (Arnold et al. 2018)

Data Sheets (Gebru et al. 2018)

Model Cards (Mitchell et al. 2018)

A “Nutrition Label” for Privacy (Kelley et al. 2009)

The Dataset Nutrition Label: A Framework To Drive Higher Data Quality Standards (Holland et al. 2019)

Data Statements for Natural Language Processing: Toward Mitigating System Bias and Enabling Better Science (Bender and Friedman 2018)

Appendix B: Diverse Voices Process and Artifacts

Procurement Recruitment Email

Procurement Confirmation Email 

Appendix C: Glossary

Sources Cited

  1. Holstein, K., Vaughan, J.W., Daumé, H., Dudík, M., & Wallach, H.M. (2018). Improving Fairness in Machine Learning Systems: What Do Industry Practitioners Need? CHI.
  2. Young, M., Magassa, L. and Friedman, B. (2019) Toward inclusive tech policy design: a method for underrepresented voices to strengthen tech policy documents. Ethics and Information Technology 21(2), 89-103.
  3. World Wide Web Consortium Process Document (W3C) process outlined here: https://www.w3.org/2019/Process-20190301/
  4. Internet Engineering Task Force (IETF) process outlined here: https://www.ietf.org/standards/process/
  5. The Web Hypertext Application Technology Working Group (WHATWG) process outlined here: https://whatwg.org/faq#process
  6. Oever, N., Moriarty, K. The Tao of IETF: A novice's guide to the Internet Engineering Task Force. https://www.ietf.org/about/participate/tao/.
  7. Young, M., Magassa, L. and Friedman, B. (2019) Toward inclusive tech policy design: a method for underrepresented voices to strengthen tech policy documents. Ethics and Information Technology 21(2), 89-103.
  8. Friedman, B, Kahn, Peter H., and Borning, A., (2008) Value sensitive design and information systems. In Kenneth Einar Himma and Herman T. Tavani (Eds.) The Handbook of Information and Computer Ethics., (pp. 70-100) John Wiley & Sons, Inc. http://jgustilo.pbworks.com/f/the-handbook-of-information-and-computer-ethics.pdf#page=104; Davis, J., and P. Nathan, L. (2015). Value sensitive design: applications, adaptations, and critiques. Handbook of Ethics, Values, and Technological Design: Sources, Theory, Values and Application Domains. (pp. 11-40) DOI: 10.1007/978-94-007-6970-0_3. https://www.researchgate.net/publication/283744306_Value_Sensitive_Design_Applications_Adaptations_and_Critiques; Borning, A. and Muller, M. (2012). Next steps for value sensitive design. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (CHI '12). (pp 1125-1134) DOI: https://doi.org/10.1145/2207676.2208560 https://dl.acm.org/citation.cfm?id=2208560
  9. Pichai, S., (2018). AI at Google: our principles. The Keyword. https://www.blog.google/technology/ai/ai-principles/; IBM’s Principles for Trust and Transparency. IBM Policy. https://www.ibm.com/blogs/policy/trust-principles/; Microsoft AI principles. Microsoft. https://www.microsoft.com/en-us/ai/our-approach-to-ai; Ethically Aligned Design – Version II. IEEE. https://standards.ieee.org/content/dam/ieee-standards/standards/web/documents/other/ead_v2.pdf
  10. Zeng, Y., Lu, E., and Huangfu, C. (2018) Linking artificial intelligence principles. CoRR https://arxiv.org/abs/1812.04814.
  11. essica Fjeld, Hannah Hilligoss, Nele Achten, Maia Levy Daniel, Sally Kagay, and Joshua Feldman, (2018). Principled artificial intelligence - a map of ethical and rights based approaches, Berkman Center for Internet and Society, https://ai-hr.cyber.harvard.edu/primp-viz.html
  12. Jobin, A., Ienca, M., & Vayena, E. (2019). Artificial Intelligence: the global landscape of ethics guidelines. arXiv preprint arXiv:1906.11668. https://arxiv.org/pdf/1906.11668.pdf
  13. Jobin, A., Ienca, M., & Vayena, E. (2019). Artificial Intelligence: the global landscape of ethics guidelines. arXiv preprint arXiv:1906.11668. https://arxiv.org/pdf/1906.11668.pdf
  14. Ananny, M., and Kate Crawford (2018). Seeing without knowing: Limitations of the transparency ideal and its application to algorithmic accountability. New Media and Society 20 (3): 973-989.
  15. Whittlestone, J., Nyrup, R., Alexandrova, A., & Cave, S. (2019, January). The Role and Limits of Principles in AI Ethics: Towards a Focus on Tensions. In Proceedings of the AAAI/ACM Conference on AI Ethics and Society, Honolulu, HI, USA (pp. 27-28). http://www.aies-conference.com/wp-content/papers/main/AIES-19_paper_188.pdf; Mittelstadt, B. (2019). AI Ethics–Too Principled to Fail? https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3391293
  16. Greene, D., Hoffmann, A. L., & Stark, L. (2019, January). Better, nicer, clearer, fairer: A critical assessment of the movement for ethical artificial intelligence and machine learning. In Proceedings of the 52nd Hawaii International Conference on System Sciences. https://scholarspace.manoa.hawaii.edu/handle/10125/59651
  17. Raji, I. D., & Buolamwini, J. (2019). Actionable auditing: Investigating the impact of publicly naming biased performance results of commercial ai products. In AAAI/ACM Conf. on AI Ethics and Society (Vol. 1). https://www.media.mit.edu/publications/actionable-auditing-investigating-the-impact-of-publicly-naming-biased-performance-results-of-commercial-ai-products/
  18. Algorithmic Impact Assessment (2019) Government of Canada https://www.canada.ca/en/government/system/digital-government/modern-emerging-technologies/responsible-use-ai/algorithmic-impact-assessment.html
  19. Benjamin, M., Gagnon, P., Rostamzadeh, N., Pal, C., Bengio, Y., & Shee, A. (2019). Towards Standardization of Data Licenses: The Montreal Data License. arXiv preprint arXiv:1903.12262. https://arxiv.org/abs/1903.12262; Responsible AI Licenses v0.1. RAIL: Responsible AI Licenses. https://www.licenses.ai/ai-licenses
  20. See Citation 5
  21. Safe Face Pledge. https://www.safefacepledge.org/; Montreal Declaration on Responsible AI. Universite de Montreal. https://www.montrealdeclaration-responsibleai.com/; The Toronto Declaration: Protecting the right to equality and non-discrimination in machine learning systems. (2018). Amnesty International and Access Now. https://www.accessnow.org/cms/assets/uploads/2018/08/The-Toronto-Declaration_ENG_08-2018.pdf ; Dagsthul Declaration on the application of machine learning and artificial intelligence for social good. https://www.dagstuhl.de/fileadmin/redaktion/Programm/Seminar/19082/Declaration/Declaration.pdf
  22. Dobbe, R., Dean, S., Gilbert, T., & Kohli, N. (2018). A Broader View on Bias in Automated Decision-Making: Reflecting on Epistemology and Dynamics. https://arxiv.org/pdf/1807.00553.pdf
  23. Wagstaff, K. (2012). Machine learning that matters. https://arxiv.org/pdf/1206.4656.pdf ; Friedman, B., Kahn, P. H., Borning, A., & Huldtgren, A. (2013). Value sensitive design and information systems. In Early engagement and new technologies: Opening up the laboratory (pp. 55-95). Springer, Dordrecht. https://vsdesign.org/publications/pdf/non-scan-vsd-and-information-systems.pdf
  24. Dobbe, R., Dean, S., Gilbert, T., & Kohli, N. (2018). A Broader View on Bias in Automated Decision-Making: Reflecting on Epistemology and Dynamics. https://arxiv.org/pdf/1807.00553.pdf
  25. Safe Face Pledge. https://www.safefacepledge.org/
  26. Montreal Declaration on Responsible AI. Universite de Montreal. https://www.montrealdeclaration-responsibleai.com/
  27. Diverse Voices How To Guide. Tech Policy Lab, University of Washington. https://techpolicylab.uw.edu/project/diverse-voices/
  28. Bender, E. M., & Friedman, B. (2018). Data statements for natural language processing: Toward mitigating system bias and enabling better science. Transactions of the Association for Computational Linguistics, 6, 587-604.
  29. Ethically Aligned Design – Version II. IEEE. https://standards.ieee.org/content/dam/ieee-standards/standards/web/documents/other/ead_v2.pdf
  30. Gebru, T., Morgenstern, J., Vecchione, B., Vaughan, J. W., Wallach, H., Daumeé III, H., & Crawford, K. (2018). Datasheets for datasets. https://arxiv.org/abs/1803.09010 https://arxiv.org/abs/1803.09010; Hazard Communication Standard: Safety Data Sheets. Occupational Safety and Health Administration, US Department of Labor. https://www.osha.gov/Publications/OSHA3514.html
  31. Holland, S., Hosny, A., Newman, S., Joseph, J., & Chmielinski, K. (2018). The dataset nutrition label: A framework to drive higher data quality standards. https://arxiv.org/abs/1805.03677; Kelley, P. G., Bresee, J., Cranor, L. F., & Reeder, R. W. (2009). A nutrition label for privacy. In Proceedings of the 5th Symposium on Usable Privacy and Security (p. 4). ACM. http://cups.cs.cmu.edu/soups/2009/proceedings/a4-kelley.pdf
  32. Mitchell, M., Wu, S., Zaldivar, A., Barnes, P., Vasserman, L., Hutchinson, B., ... & Gebru, T. (2019, January). Model cards for model reporting. In Proceedings of the Conference on Fairness, Accountability, and Transparency (pp. 220-229). ACM. https://arxiv.org/abs/1810.03993
  33. Hind, M., Mehta, S., Mojsilovic, A., Nair, R., Ramamurthy, K. N., Olteanu, A., & Varshney, K. R. (2018). Increasing Trust in AI Services through Supplier's Declarations of Conformity. https://arxiv.org/abs/1808.07261
  34. Veale M., Van Kleek M., & Binns R. (2018) ‘Fairness and Accountability Design Needs for Algorithmic Support in High-Stakes Public Sector Decision-Making’ in Proceedings of the ACM Conference on Human Factors in Computing Systems, CHI 2018. https://arxiv.org/abs/1802.01029.
  35. Benjamin, M., Gagnon, P., Rostamzadeh, N., Pal, C., Bengio, Y., & Shee, A. (2019). Towards Standardization of Data Licenses: The Montreal Data License. https://arxiv.org/abs/1903.12262
  36. Cooper, D. M. (2013, April). A Licensing Approach to Regulation of Open Robotics. In Paper for presentation for We Robot: Getting down to business conference, Stanford Law School.
  37. Responsible AI Practices. Google AI. https://ai.google/education/responsible-ai-practices
  38. Everyday Ethics for Artificial Intelligence. (2019). IBM. https://www.ibm.com/watson/assets/duo/pdf/everydayethics.pdf
  39. Federal Trade Commission. (2012). Best Practices for Common Uses of Facial Recognition Technologies (Staff Report). Federal Trade Commission, 30. https://www.ftc.gov/sites/default/files/documents/reports/facing-facts-best-practices-common-uses-facial-recognition-technologies/121022facialtechrpt.pdf
  40. Microsoft (2018). Responsible bots: 10 guidelines for developers of conversational AI. https://www.microsoft.com/en-us/research/uploads/prod/2018/11/Bot_Guidelines_Nov_2018.pdf
  41. Tramer, F., Atlidakis, V., Geambasu, R., Hsu, D., Hubaux, J. P., Humbert, M., ... & Lin, H. (2017, April). FairTest: Discovering unwarranted associations in data-driven applications. In 2017 IEEE European Symposium on Security and Privacy (EuroS&P) (pp. 401-416). IEEE. https://github.com/columbia/fairtest, https://www.mhumbert.com/publications/eurosp17.pdf
  42. Kishore Durg (2018). Testing AI: Teach and Test to raise responsible AI. Accenture Technology Blog. https://www.accenture.com/us-en/insights/technology/testing-AI
  43. Kush R. Varshney (2018). Introducing AI Fairness 360. IBM Research Blog. https://www.ibm.com/blogs/research/2018/09/ai-fairness-360/
  44. Dave Gershgorn (2018). Facebook says it has a tool to detect bias in its artificial intelligence. Quartz. https://qz.com/1268520/facebook-says-it-has-a-tool-to-detect-bias-in-its-artificial-intelligence/
  45. James Wexler. (2018) The What-If Tool: Code-Free Probing of Machine Learning Models. Google AI Blog. https://ai.googleblog.com/2018/09/the-what-if-tool-code-free-probing-of.html
  46. Miro Dudík, John Langford, Hanna Wallach, and Alekh Agarwal (2018). Machine Learning for fair decisions. Microsoft Research Blog. https://www.microsoft.com/en-us/research/blog/machine-learning-for-fair-decisions/
  47. Veale, M., Binns, R., & Edwards, L. (2018). Algorithms that Remember: Model Inversion Attacks and Data Protection Law. Phil. Trans. R. Soc. A, 376, 20180083. https://doi.org/10/gfc63m
  48. Floridi, L. (2010, February). Information: A Very Short Introduction.
  49. Data Information Specialists Committee UK, 2007. http://www.disc-uk.org/qanda.html.
  50. Harwell, Drew. “Federal Study Confirms Racial Bias of Many Facial-Recognition Systems, Casts Doubt on Their Expanding Use.” The Washington Post, WP Company, 21 Dec. 2019, www.washingtonpost.com/technology/2019/12/19/federal-study-confirms-racial-bias-many-facial-recognition-systems-casts-doubt-their-expanding-use/
  51. Hildebrandt, M. (2019) ‘Privacy as Protection of the Incomputable Self: From Agnostic to Agonistic Machine Learning’, Theoretical Inquiries in Law, 20(1) 83–121.
  52. D'Amour, A., Heller, K., Moldovan, D., Adlam, B., Alipanahi, B., Beutel, A., ... & Sculley, D. (2020). Underspecification presents challenges for credibility in modern machine learning. arXiv preprint arXiv:2011.03395.
  53. Selinger, E. (2019). ‘Why You Can’t Really Consent to Facebook’s Facial Recognition’, One Zero. https://onezero.medium.com/why-you-cant-really-consent-to-facebook-s-facial-recognition-6bb94ea1dc8f
  54. Lum, K., & Isaac, W. (2016). To predict and serve?. Significance, 13(5), 14-19. https://rss.onlinelibrary.wiley.com/doi/full/10.1111/j.1740-9713.2016.00960.x
  55. LabelInsight (2016). “Drive Long-Term Trust & Loyalty Through Transparency”. https://www.labelinsight.com/Transparency-ROI-Study
  56. Crawford and Paglen, https://www.excavating.ai/
  57. Geva, Mor & Goldberg, Yoav & Berant, Jonathan. (2019). Are We Modeling the Task or the Annotator? An Investigation of Annotator Bias in Natural Language Understanding Datasets. https://arxiv.org/pdf/1908.07898.pdf
  58. Bender, E. M., & Friedman, B. (2018). Data statements for natural language processing: Toward mitigating system bias and enabling better science. Transactions of the Association for Computational Linguistics, 6, 587-604.
  59. Desmond U. Patton et al (2017).
  60. See Cynthia Dwork et al.,
  61. Katta Spiel, Oliver L. Haimson, and Danielle Lottridge. (2019). How to do better with gender on surveys: a guide for HCI researchers. Interactions. 26, 4 (June 2019), 62-65. DOI: https://doi.org/10.1145/3338283
  62. A. Doan, A. Y. Halevy, and Z. G. Ives. Principles of Data Integration. Morgan Kaufmann, 2012
  63. Momin M. Malik. (2019). Can algorithms themselves be biased? Medium. https://medium.com/berkman-klein-center/can-algorithms-themselves-be-biased-cffecbf2302c
  64. Fire, Michael, and Carlos Guestrin (2019). “Over-Optimization of Academic Publishing Metrics: Observing Goodhart’s Law in Action.” GigaScience 8 (giz053). https://doi.org/10.1093/gigascience/giz053.
  65. Vogelsang, A., & Borg, M. (2019, September). Requirements engineering for machine learning: Perspectives from data scientists. In 2019 IEEE 27th International Requirements Engineering Conference Workshops (REW) (pp. 245-251). IEEE
  66. Eckersley, P. (2018). Impossibility and Uncertainty Theorems in AI Value Alignment (or why your AGI should not have a utility function). arXiv preprint arXiv:1901.00064.
  67. Partnership on AI. Report on Algorithmic Risk Assessment Tools in the U.S. Criminal Justice System, Requirement 5.
  68. Eckersley, P. (2018). Impossibility and Uncertainty Theorems in AI Value Alignment (or why your AGI should not have a utility function). arXiv preprint arXiv:1901.00064.https://arxiv.org/abs/1901.00064
  69. If it is not, there is likely a bug in the code. Checking a predictive model's performance on the training set cannot distinguish irreducible error (which comes from intrinsic variance of the system) from error introduced by bias and variance in the estimator; this is universal, and has nothing to do with different settings or
  70. Selbst, Andrew D. and Boyd, Danah and Friedler, Sorelle and Venkatasubramanian, Suresh and Vertesi, Janet (2018). “Fairness and Abstraction in Sociotechnical Systems”, ACM Conference on Fairness, Accountability, and Transparency (FAT*). https://ssrn.com/abstract=3265913
  71. Tools that can be used to explore and audit the predictive model fairness include FairML, Lime, IBM AI Fairness 360, SHAP, Google What-If Tool, and many others
  72. Wagstaff, K. (2012). Machine learning that matters. arXiv preprint arXiv:1206.4656. https://arxiv.org/abs/1206.4656
Table of Contents
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16