Information Technology Reference
In-Depth Information
Fig. 3 Reliability versus time
showing error
Fig. 4 Reliability versus time
without error
occurrence rate drastically increases, following which the failure rate decreases
gradually as failures are detected and
fixed. Each upward slope of the graph rep-
resents an upgrade.
5.2 Results of Support Vector Algorithm
As already mentioned in Sect.
3
, an interval-domain data set representing the
cumulative number of failure per week is used here for the purpose of software
reliability analysis. Support Vector Regression is the machine learning technique
used for the same which, given the number of failures that occur in the current
week, predicts the expected number of failures for the next week. The training data
Search WWH ::
Custom Search