Defect classification was performed by association rule mining and artificial neural network. Association rule
mining algorithm occasionally contributes to useless rules. Therefore it is really problematic for flaw classification
predicated on these useless rules. To be able to prevent such problems, we've improved the guidelines applying
Adaptive PSO algorithm before classification predicated on help and confidence value. In that report we've labeled
the problems applying artificial neural network. The principles were removed from the input applying association
rule mining. The principles were improved applying adaptive PSO algorithm. Problems were labeled applying
artificial neural network. The product quality was sure utilizing the quality metrics such as for instance flaw density,
Sensitivity etc. Our planned process has accomplished a precision value of 99.5%. It's larger in comparison with the
prevailing method. Hence the efficiency methods formula revealed which our planned process is efficient.