Machine Learning Methods for Predicting Failures in Hard Drives: A Multiple-Instance Application

By Joseph F. Murray, Gordon F. Hughes, Kenneth Kreutz-Delgado. Journal of machine learning research, 2005

305 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/562a596836f42714d81e1f861671959ba12e0246

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Multiple instance ranking (2008, methodology): “Other applications of MIC include automatic image annotation (Andrews et al., 2003), context-based image indexing (Maron & Ratan, 1998), text categorization (Andrews et al., 2003) and hard-drive failure prediction (Murray et al., 2005).”
An Extension on "Statistical Comparisons of Classifiers over Multiple Data Sets" for all Pairwise Comparisons (2008, background): “Although the employment of many data sets could seem beneficial in order to improve the generalization of results, in some specific domains, that is, imbalanced classification (Owen, 2007) or multi-instance classification (Murray et al., 2005), data sets are difficult to find.”
A process for predicting manhole events in Manhattan (2010, background): “Note that many other problems dealing with prediction of rare events in continuous time rely on a shorter time scale, including seizure prediction for epileptic patients, or prediction of failures in hard drives (Murray et al. 2005).”
Identifying Multi-instance Outliers (2010, unknown): “Current hard-drive reliability inspection is studied through multi-instance learning [2].”
Fast Bundle Algorithm for Multiple-Instance Learning (2012, methodology): “Computational results show this method is linearly scalable, while not sacrificing generalization accuracy, permitting modeling on new and larger data sets in computational chemistry and other applications.”
Multiple instance learning: A survey of problem characteristics and applications (2016, methodology): “A similar model is used for the prediction of hard drive failure [161].”
On classification with bags, groups and sets (2014, unknown): “Other applications where the bag of instances representation has been used are detecting hard drive failures [40], detecting fraudulent financial accounts [21], music information retrieval [34], and spam filtering and advertising [44].”