Machine learning is increasingly being used to predict individuals’ attitudes, behaviors, and preferences across an array of applications — from personalized marketing to precision medicine. Unsurprisingly, given the speed of change and ever-increasing complexity, there have been several recent high-profile examples of “machine learning gone wrong.”
Make “Fairness by Design” Part of Machine Learning
Five ways to keep bias out of your models.
August 01, 2018
Summary.
Bias in machine learning is a real problem. When models don’t perform as intended, people and process are normally to blame. But it’s possible to employ a “fairness by design” strategy to machine learning, encompassing a few key facets. To do so, companies can take the following steps: pair data scientists with a social scientist; annotate with caution; couple traditional machine learning metrics with fairness measures; when sampling, balance representativeness with critical mass constraints; and keep de-biasing in mind when building models.
New!
HBR Learning
Digital Intelligence Course
Accelerate your career with Harvard ManageMentor®. HBR Learning’s online leadership training helps you hone your skills with courses like Digital Intelligence . Earn badges to share on LinkedIn and your resume. Access more than 40 courses trusted by Fortune 500 companies.
Excel in a world that's being continually transformed by technology.
Learn More & See All Courses
For HBR Subscribers
AI, Algorithms, and Bias
Algorithms can improve our predictions and decisions, but they can also perpetuate our blind spots and biases. Here’s what you need to know about the problem—and how organizations can address it.
Show Reading List
New!
HBR Learning
Digital Intelligence Course
Accelerate your career with Harvard ManageMentor®. HBR Learning’s online leadership training helps you hone your skills with courses like Digital Intelligence . Earn badges to share on LinkedIn and your resume. Access more than 40 courses trusted by Fortune 500 companies.
Excel in a world that's being continually transformed by technology.