Using computational learning theory to analyze the testing complexity of program segments

Abstract

We examine the complexity of testing different program constructs by defining a measure called VCP-dimension and applying it to classes of programs, where all programs in a class share the same syntactic structure. VCP-dimension gives bounds on the number of test points needed to determine approximate correctness, so it gives insight into the difficulty of testing a program construct represented by a program class. We investigate the VCP-dimension of straight line code, if-then-else statements and for loops, and we compare the VCP-dimension of different combinations of constructs.<>