Malware datasets merge and implement feature selection method

The project is to merge dataset or preprocess them to prepare for train and test dataset on machine learning which are the different types families of malware. The aim is to merge two feature selection methods (for example n-gram and Genetic algorithms) to select the relevant ones and discuss the result. Objective is to find a new or enhance method for feature selection which reduce the issues of feature selection such as high diminsitnolity. The dataset used in the below link paper (i had the dataset): [login to view URL]