Every sample is repeated？

CGCL-codes / VulDeePecker

VulDeePecker: A Deep Learning-Based System for Vulnerability Detection

Apache License 2.0

293 stars 103 forks source link

Every sample is repeated？ #9

Open anmilky opened 5 years ago

anmilky commented 5 years ago

hi, may u explain the reason why every sample is repeated？ i saw every sample is repeated and each sample is marked twice, cfunc and cppfunc respectively。 may i ask you why? Are these data separated during training?

VulDeePecker commented 5 years ago

These code gadgets are generated by different rules by Checkmarx. The training programs are separated from the target programs, that is, the programs for training are not used as the target programs for testing.