The What-If Tool: Interactive Probing of Machine Learning Models

Wexler, James; Pushkarna, Mahima; Bolukbasi, Tolga; Wattenberg, Martin; Viégas, Fernanda B.; Wilson, Jimbo

doi:10.1109/tvcg.2019.2934619

Cited by 311 publications

(292 citation statements)

References 21 publications

Supporting

Mentioning

261

Contrasting

Unclassified

Order By: Relevance

“…In contrast, we provide a range of views that are linked through subsets of the classification data and allow in‐depth comparisons between models. The What‐If‐Tool [WPB*20] enables users to compose a range of visualizations, including bar chars and confusion matrices on subsets of their data by slicing sets based on feature values. In contrast, we enable the creation of more complex subsets based on set algebra, and enable comparisons between models based on these sets.…”

Section: Related Workmentioning

confidence: 99%

Boxer: Interactive Comparison of Classifier Results

Gleicher

Barve

et al. 2020

Computer Graphics Forum

View full text Add to dashboard Cite

Machine learning practitioners often compare the results of different classifiers to help select, diagnose and tune models. We present Boxer, a system to enable such comparison. Our system facilitates interactive exploration of the experimental results obtained by applying multiple classifiers to a common set of model inputs. The approach focuses on allowing the user to identify interesting subsets of training and testing instances and comparing performance of the classifiers on these subsets. The system couples standard visual designs with set algebra interactions and comparative elements. This allows the user to compose and coordinate views to specify subsets and assess classifier performance on them. The flexibility of these compositions allow the user to address a wide range of scenarios in developing and assessing classifiers. We demonstrate Boxer in use cases including model selection, tuning, fairness assessment, and data quality diagnosis.

show abstract

Section: Related Workmentioning

confidence: 99%

Boxer: Interactive Comparison of Classifier Results

Gleicher

Barve

et al. 2020

Computer Graphics Forum

View full text Add to dashboard Cite

show abstract

“…There has been development of semi-automated tools to help practitioners detect subgroup biases [2,3,4,7,9,20,21,22]. However, these tools are often designed in isolation from users [18].…”

Section: Background and Related Workmentioning

confidence: 99%

Designing Tools for Semi-Automated Detection of Machine Learning Biases: An Interview Study

Law

Malik

et al. 2020

Preprint

View full text Add to dashboard Cite

1 Machine learning models often make predictions that bias against certain subgroups of input data. When undetected, machine learning biases can constitute significant financial and ethical implications. Semi-automated tools that involve humans in the loop could facilitate bias detection. Yet, little is known about the considerations involved in their design. In this paper, we report on an interview study with 11 machine learning practitioners for investigating the needs surrounding semi-automated bias detection tools. Based on the findings, we highlight four considerations in designing to guide system designers who aim to create future tools for bias detection.

show abstract

“…To address these types of issues, several approaches exist: for instance, researchers have recently released a tool called 'What-If', an open-source application that lets practitioners not only visualize their data, but also test the performance of their ML model in hypothetical situations, for instance modifying some characteristics of data points and analyzing subsequent model behavior, by measuring fairness metrics such as Equal Opportunity and Demographic Parity [Wexler et al, 2019]. Other approaches address bias by changing the training procedure or the structure of ML models themselves, for instance by transforming the raw data in a space in which discriminatory information cannot be found [Zemel et al, 2013] or using a variational autoencoder to learn the latent structure from the dataset and using this structure to re-weight the importance of specific data points during model training [Ribeiro et al, 2016].…”

Section: Numerical Biasmentioning

confidence: 99%

On the Morality of Artificial Intelligence [Commentary]

Luccioni

Bengio²

2020

IEEE Technol. Soc. Mag.

View full text Add to dashboard Cite

Much of the existing research on the social and ethical impact of Artificial Intelligence has been focused on defining ethical principles and guidelines surrounding Machine Learning (ML) and other Artificial Intelligence (AI) algorithms [IEEE, 2017, Jobin et al., 2019. While this is extremely useful for helping define the appropriate social norms of AI, we believe that it is equally important to discuss both the potential and risks of ML and to inspire the community to use ML for beneficial objectives. In the present article, which is specifically aimed at ML practitioners, we thus focus more on the latter, carrying out an overview of existing high-level ethical frameworks and guidelines, but above all proposing both conceptual and practical principles and guidelines for ML research and deployment, insisting on concrete actions that can be taken by practitioners to pursue a more ethical and moral practice of ML aimed at using AI for social good. * Also CIFAR Senior Fellow

show abstract

The What-If Tool: Interactive Probing of Machine Learning Models

Cited by 311 publications

References 21 publications

Boxer: Interactive Comparison of Classifier Results

Boxer: Interactive Comparison of Classifier Results

Designing Tools for Semi-Automated Detection of Machine Learning Biases: An Interview Study

On the Morality of Artificial Intelligence [Commentary]

Contact Info

Product

Resources

About