Fig 1.
Proposed Methodology: (a) Shows Phase I, where Parallel and Distributed Processing is Conducted; (b) Shows Phase II, which Includes the Training of Deep Learning Model Using None-faulty Data from Phase I, (c) Shows the Phase III, where the Output of the Faulty Processor is Produced Using Deep Learning Model.
Fig 2.
Example of missing data in human gait dataset [50].
Fig 3.
Some examples of missing data in driver distraction dataset [51].
Table 1.
Datasets description.
Fig 4.
Driver distraction dataset results using four deep learning models.
Table 2.
Driver distraction results.
Fig 5.
Gait human dataset results using four deep learning models.
Table 3.
Human gait detection results.
Fig 6.
Vulnerability dataset results using four deep learning models.
Fig 7.
Best vulnerability dataset results based on loss function.
Fig 8.
Best missing vulnerability dataset results based on loss function.
Fig 9.
KDD Cup99 dataset results using four deep learning models.
Fig 10.
Best KDD Cup99 dataset results based on loss function.
Fig 11.
Best missing KDD Cup99 dataset results based on loss function.