NORMA eResearch @NCI Library

Early Prediction of HDD Failures in the Cloud Using Interpretable AI Models

Patil, Ashwini (2022) Early Prediction of HDD Failures in the Cloud Using Interpretable AI Models. Masters thesis, Dublin, National College of Ireland.

[thumbnail of Master of Science]
PDF (Master of Science)
Download (1MB) | Preview
[thumbnail of Configuration manual]
PDF (Configuration manual)
Download (765kB) | Preview


Hard Disk Drive(HDD) storage devices are essential to the cloud datacenter’s dependability. The availability of cloud services can be significantly impacted by HDD failures, which can be expensive for both the cloud provider and the end user financially and in terms of reputation. To reduce the effects of failure, it is crucial to foresee HDD failure beforehand and arrange the necessary corrective actions. In this study, we offer an HDD failure prediction system that uses RUL(Remaining Useful Life) predictions to anticipate HDD failures in advance. To accurately forecast failures and enable simple user understanding of prediction findings, the suggested system uses a composite model composed of bi-directional LSTM(Long Short-Term Memory) and RFC(Random Forest Classification)/OCT(Optimal Classification Tree). The models are trained and evaluated using publicly available Backblaze dataset. The model performance of composite model was evaluated against the baseline RFC and OCT models. From the experiments, we observed the composite model made up of RFC and bi-directional LSTM had an accuracy of 99% against the baseline RFC model accuracy of 94%. This promising result requires further experimental validation.

Item Type: Thesis (Masters)
Subjects: Q Science > QA Mathematics > Electronic computers. Computer science
T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science
Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence
Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence
T Technology > T Technology (General) > Information Technology > Cloud computing
Divisions: School of Computing > Master of Science in Cloud Computing
Depositing User: Tamara Malone
Date Deposited: 08 Dec 2022 11:55
Last Modified: 08 Mar 2023 14:26

Actions (login required)

View Item View Item