Access your Pro+ Content below.
Big data throws big biases into machine learning data sets
This article is part of the Business Information issue of February 2018, Vol. 6 No. 1
Say you're training an image recognition system to identify U.S. presidents. The historical data reveals a pattern of males, so the algorithm concludes that only men are presidents. It won't recognize a female in that role, even though it's a probable outcome in future elections. This latent bias is one of the many types of biases that challenge data scientists today. If the machine learning data set they use in an AI project isn't neutral -- and it's safe to say almost no data is -- the outcomes can actually amplify bias and discrimination that's present in the machine learning data set. Visual recognition technologies that label images require vast amounts of labeled data, which largely comes from the web. You can imagine the dangers in that -- and researchers at the University of Washington and University of Virginia confirmed one poignant example of gender bias in a recent report. They found that when a visual semantic role labeling system sees a spatula, it labels the utensil as a cooking tool, but it's also likely to refer...
Access this PRO+ Content for Free!
By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
Features in this issue
AI holds massive potential for good, but it also amplifies negative outcomes if data scientists don't recognize data biases and correct them in machine learning data sets.
Limited AI capabilities could soon give way to technology that is truly transformative for enterprises, surpassing the overhyped functionality that we see today.
Artificial intelligence is making inroads into manufacturing systems. Case in point is printed circuits manufacturer Jabil, which uses AI to improve its processes and products.
Providers and big vendors are wary, but healthcare in the cloud and the benefits artificial intelligence provides will spur a shift from on-premises systems for analyzing EHRs.
Columns in this issue
Machine learning and other artificial intelligence technologies are poised to offer businesses big benefits, but companies have to walk before they run with AI.
Cognitive computing and healthcare data aggregation may prove to be important bedside companions to doctors treating their patients in the era of medical specialization.
Amid the infiltration of artificial intelligence into business and society, spawning fears from some of Armageddon, humankind's greatest AI threat may very well come from within.