Srakaew et al. [18]

Data Mining Classifier Algorithm i.e. J48 by generating decision trees.

Reference Data Set: 1200 files in total out of which 900 are malicious and 300 are non-malicious.

Application Data Set: 3251 files in total out of which 2951 are malicious and 300 are non-malicious.

Reference Data Set:

Allapple, Podhuha and Virut viral families each containing 300 malicious samples.

Application Data Set:

Allapple, Podhuha and Virut viral families with 890, 8 and 2,053 malicious samples, respectively.

Statistical Features Approach:

Conversion of malicious and non-malicious files into hex sequences for extracting statistical aspects using n-grams of bytes.

Abstract Assembly Approach:

Conversion of malicious and non-malicious files into assembly instructions for extracting selected instructions using n-grams of interesting opcodes.