Measuring Software Reliability: A Trend Using Machine Learning Techniques - Complex System Modelling and Control Through Intelligent Soft Computations - page 808

Fig. 11 SVM classi cation
plot with real values

Fig. 12 Regression plot
based decision

initial step to foster further extension under different software metrics 3 (refer
Fig. 12 ). Further, for both SVM real values and regression based real values, the
regression based ML slightly out performs in terms of precision and therefore can
be further chosen as standard model. For this present study data set was restricted
and thus given python code can only be extended towards few metrics mentioned
instead testing all the other statistical measures suggested (Figs. 11 and 13 ).
The entire simulation and validation is performed in restricted environment
under open source NetWorkX 4 environment supported by Python Language data
structures.

3 LOC: Lines of Code (not a C&K metric), DAM: Data Access Metric (QMOOD metric suite)
MOA: Measure of Aggregation (QMOOD metric suite), MFA: Measure of Functional Abstraction
(QMOOD metric suite), CAM: Cohesion Among Methods of Class (QMOOD metric suite), IC:
Inheritance Coupling (quality oriented extension for the C&K metric suite), CBM: Coupling
Between Methods (quality oriented extension for the C&K metric suite), AMC: Average Method
Complexity (quality oriented extension to C&K metric suite), CC: McCabe's Cyclomatic
Complexity.
4

https://networkx.github.io/ .