Detecting Malware with an Ensemble Method Based on Deep Neural Network

Algorithm 1

Opcode sequence extraction algorithm for executive files.

Input: Executive file

Output: Opcode sequence

   = ; // Get all executive files;

for in;

= ; // Open the corresponding IDA pro decompiled file;

for in; // Read in line;

= t(“  ”); // Cut the line into phrases by space character;

for    in  ;

//To judge each phrase, it requires to meet the following two points at the same time:

The current word belongs to opcode set opcode_set;

The last three words are not duplicated opcodes.

if    in    and   =   and   =

= ;

= ;

;

(11)end if

(12) end for

(13)end for

(14) end for