Using Machine Learning to Predict Hard Drive Failures

Logo
Presented by

Daisy Zhuo, Interpretable AI; Andy Klein, Backblaze

About this talk

Daisy Zhou and her colleagues at Interpretable AI recently published a paper on how they have used the Machine Learning techniques to predict hard drive failure. Their analysis and process delivers useful insights into drive failure even with limited historical data, enabling organizations to make replacement decisions even when data collection has only recently begun. Unlike current black-box methods of drive failure prediction, the application of interpretable machine learning methods is transparent and completely understandable by humans. So much so, that previous knowledge of Machine Learning techniques is not required to enjoy and benefit from this presentation.

Related topics:

More from this channel

Upcoming talks (1)
On-demand talks (66)
Subscribers (10012)
For over a decade, Backblaze has been a transparent company in the opaque world of cloud data storage. We've written on our blog about hard drive failure rates, data center internals, our open source high density storage design, and more. Here we are on Brighttalk and each month our goal is to bring you a webcast or video that sheds some light behind the scenes of a Cloud Storage company. And yes, we'll have a sales pitch or two the "off" weeks. Join us when you can!