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 (0)
On-demand talks (107)
Subscribers (20379)
Backblaze is a leading independent cloud provider that makes it astonishingly easy for people to store, protect, and use their data. Its B2 Cloud Storage platform offers always-hot, S3 compatible object storage that’s readily available through APIs, CLI, Web UI, and 3rd party integrations--to seamlessly support ranging workflows and fit hybrid cloud, multi-cloud, and other IaaS strategies. Its computer backup solutions offer automatic data protection for business fleets and home Mac and PC. We're here on BrightTalk to unpack product and solution updates, discuss best practices around data storage and protection, and openly share hard drive stats based on our experience spinning >200,000 drives. Thanks for joining us.