Comparative and Analysis Study for Malicious Executable by Using Various Classification Algorithms

摘要

There are a lot of applications regarding the data mining methods in detecting malwares. One of the most widely utilized data mining methods is the Classification method. In our research, we are presenting a data mining classification procedure through applying machine learning algorithms to detect malicious executable files, and this study will investigate the approach of classification in some algorithms such as (Support Vector Machine. Random Forest. KNN (k-Nearest Neighbors Classifier), and The Hoeffding Tree). In our classification process, we used some of well-known machine-learning algorithms by WEKA libraries, and then we train our dataset to detect malware. We made a comparative analysis between algorithms used and how they deal with the selected features based on the size of the data, to illustrate the performance efficiency Where we go t a high accuracy up to 98% with Random Forest. Moreover, this study is considered as a base for future studies regarding malware analysis through machine learning algorithms.