- A new technique compares the reasoning of a machine-learning model to that of a human.
- Called Shared Interest, the strategy incorporates quantifiable metrics that compare how well reasoning matches.
- This could help a user easily uncover concerning trends in a model’s decision-making.
- For example, perhaps the model often becomes confused by distracting, irrelevant features, like background objects in photos.
In machine learning, understanding why a model makes certain decisions is often just as important as whether those decisions are correct. For instance, a machine-learning model might correctly predict that a skin lesion is cancerous, but it could have done so using an unrelated blip on a clinical photo.
While tools exist to help experts make sense of a model’s reasoning, often these methods only provide insights on one decision at a time, and each must be manually evaluated. Models are commonly trained using millions of data inputs, making it almost impossible for a human to evaluate enough decisions to identify patterns.
Now, researchers at MIT and IBM Research have created a method that enables a user to aggregate, sort, and rank these individual explanations to rapidly analyze a machine-learning model’s behavior. Their technique, called Shared Interest, incorporates quantifiable metrics that compare how well a model’s reasoning matches that of a human.
Shared Interest could help a user easily uncover concerning trends in a model’s decision-making — for example, perhaps the model often becomes confused by distracting, irrelevant features, like background objects in photos. Aggregating these insights could help the user quickly and quantitatively determine whether a model is trustworthy and ready to be deployed in a real-world situation.
“In developing Shared Interest, our goal is to be able to scale up this analysis process so that you could understand on a more global level what your model’s behavior is,” says lead author Angie Boggust, a graduate student in the Visualization Group of the Computer Science and Artificial Intelligence Laboratory (CSAIL).
Boggust wrote the paper with her advisor, Arvind Satyanarayan, an assistant professor of computer science who leads the Visualization Group, as well as Benjamin Hoover and senior author Hendrik Strobelt, both of IBM Research. The paper will be presented at the Conference on Human Factors in Computing Systems.
Boggust began working on this project during a summer internship at IBM, under the mentorship of Strobelt. After returning to MIT, Boggust and Satyanarayan expanded on the project and continued the collaboration with Strobelt and Hoover, who helped deploy the case studies that show how the technique could be used in practice.
Human-artificial intelligence alignment
Shared Interest leverages popular techniques that show how a machine-learning model made a specific decision, known as saliency methods. If the model is classifying images, saliency methods highlight areas of an image that are important to the model when it made its decision. These areas are visualized as a type of heatmap, called a saliency map, that is often overlaid on the original image. If the model classified the image as a dog, and the dog’s head is highlighted, that means those pixels were important to the model when it decided the image contains a dog.
Shared Interest works by comparing saliency methods to ground-truth data. In an image dataset, ground-truth data are typically human-generated annotations that surround the relevant parts of each image. In the previous example, the box would surround the entire dog in the photo. When evaluating an image classification model, Shared Interest compares the model-generated saliency data and the human-generated ground-truth data for the same image to see how well they align.
The technique uses several metrics to quantify that alignment (or misalignment) and then sorts a particular decision into one of eight categories. The categories run the gamut from perfectly human-aligned (the model makes a correct prediction and the highlighted area in the saliency map is identical to the human-generated box) to completely distracted (the model makes an incorrect prediction and does not use any image features found in the human-generated box).
“On one end of the spectrum, your model made the decision for the exact same reason a human did, and on the other end of the spectrum, your model and the human are making this decision for totally different reasons. By quantifying that for all the images in your dataset, you can use that quantification to sort through them,” Boggust explains.
The technique works similarly with text-based data, where key words are highlighted instead of image regions.
Have you read?
The researchers used three case studies to show how Shared Interest could be useful to both nonexperts and machine-learning researchers.
In the first case study, they used Shared Interest to help a dermatologist determine if he should trust a machine-learning model designed to help diagnose cancer from photos of skin lesions. Shared Interest enabled the dermatologist to quickly see examples of the model’s correct and incorrect predictions. Ultimately, the dermatologist decided he could not trust the model because it made too many predictions based on image artifacts, rather than actual lesions.
“The value here is that using Shared Interest, we are able to see these patterns emerge in our model’s behavior. In about half an hour, the dermatologist was able to make a confident decision of whether or not to trust the model and whether or not to deploy it,” Boggust says.
In the second case study, they worked with a machine-learning researcher to show how Shared Interest can evaluate a particular saliency method by revealing previously unknown pitfalls in the model. Their technique enabled the researcher to analyze thousands of correct and incorrect decisions in a fraction of the time required by typical manual methods.
In the third case study, they used Shared Interest to dive deeper into a specific image classification example. By manipulating the ground-truth area of the image, they were able to conduct a what-if analysis to see which image features were most important for particular predictions.
The researchers were impressed by how well Shared Interest performed in these case studies, but Boggust cautions that the technique is only as good as the saliency methods it is based upon. If those techniques contain bias or are inaccurate, then Shared Interest will inherit those limitations.
In the future, the researchers want to apply Shared Interest to different types of data, particularly tabular data which is used in medical records. They also want to use Shared Interest to help improve current saliency techniques. Boggust hopes this research inspires more work that seeks to quantify machine-learning model behavior in ways that make sense to humans.
This work is funded, in part, by the MIT-IBM Watson AI Lab, the United States Air Force Research Laboratory, and the United States Air Force Artificial Intelligence Accelerator.
How is the World Economic Forum ensuring that artificial intelligence is developed to benefit all stakeholders?
Artificial intelligence (AI) is impacting all aspects of society — homes, businesses, schools and even public spaces. But as the technology rapidly advances, multistakeholder collaboration is required to optimize accountability, transparency, privacy and impartiality.
The World Economic Forum's Platform for Shaping the Future of Technology Governance: Artificial Intelligence and Machine Learning is bringing together diverse perspectives to drive innovation and create trust.
- One area of work that is well-positioned to take advantage of AI is Human Resources — including hiring, retaining talent, training, benefits and employee satisfaction. The Forum has created a toolkit Human-Centred Artificial Intelligence for Human Resources to promote positive and ethical human-centred use of AI for organizations, workers and society.
- Children and young people today grow up in an increasingly digital age in which technology pervades every aspect of their lives. From robotic toys and social media to the classroom and home, AI is part of life. By developing AI standards for children, the Forum is working with a range of stakeholders to create actionable guidelines to educate, empower and protect children and youth in the age of AI.
- The potential dangers of AI could also impact wider society. To mitigate the risks, the Forum is bringing together over 100 companies, governments, civil society organizations and academic institutions in the Global AI Action Alliance to accelerate the adoption of responsible AI in the global public interest.
- AI is one of the most important technologies for business. To ensure C-suite executives understand its possibilities and risks, the Forum created the Empowering AI Leadership: AI C-Suite Toolkit, which provides practical tools to help them comprehend AI’s impact on their roles and make informed decisions on AI strategy, projects and implementations.
- Shaping the way AI is integrated into procurement processes in the public sector will help define best practice which can be applied throughout the private sector. The Forum has created a set of recommendations designed to encourage wide adoption, which will evolve with insights from a range of trials.
- The Centre for the Fourth Industrial Revolution Rwanda worked with the Ministry of Information, Communication Technology and Innovation to promote the adoption of new technologies in the country, driving innovation on data policy and AI – particularly in healthcare.
Contact us for more information on how to get involved.