IBM launches AI bias tool

Donna Miller
September 21, 2018

"It's created to translate algorithmic research from the lab into the actual practice of domains as wide-ranging as finance, human capital management, healthcare, and education", they outlined.

By releasing AI Fairness 360 as a companion open source project to the Adversarial Robustness Toolbox, IBM hopes to encourage the contributions of researchers outside the organization, and eventually build a community dedicated to improving algorithms, according to a blog post.

The Kit is supposed to analyze how the machine-learning algorithms make their decisions in real-time and work out if and how they are being biased on accident, by, for example, not identifying correctly people who are not white in photographs.

Importantly, it also automatically recommends data to add to the model to help mitigate any bias it has detected. The explanation will include the factors that resulted in a particular bias decision.

Cubs’ Addison Russell accused of physical, verbal abuse by ex-wife
The league said Reidy had declined to cooperate with investigators a year ago , but new efforts would be made. A week after flying home, I finally made the call and took legal action to start our separation".

Besides detecting automated decisions, it will also keep track of the accuracy, fairness, performance, and lineage of the system.

All of this information is presented in a visual dashboard which means a business can access this information at a glance.

IBM launched a new service to give businesses transparency into how artificial intelligence makes decisions, in a move the company says is a "major step in breaking open the black box of AI". While other open-source resources have focused exclusively on checking for bias in training data, the IBM AI Fairness 360 toolkit created by IBM Research will help check for and mitigate bias in AI models.

In addition, IBM Research will also release into the open source community an AI bias detection and mitigation toolkit that will help developers design features to address bias in AI. "AIF360 is a bit different from now available open source efforts1 due its focus on bias mitigation (as opposed to simply on metrics), its focus on industrial usability, and its software engineering", wrote Kush Varshney, principal research staff member and manager at IBM Research.

Other reports by

Discuss This Article

FOLLOW OUR NEWSPAPER