The Genetic Algorithm (GA) is a branch of evolutionary algorithms that has been proving power and success as astrategy to optimize various problems, but they may suffer fromthe slow and premature convergence of results, impacting thefinal efficiency obtained. In this perspective, variants of the GAare being developed, seeking to obtain greater convergence andprecision of results. This work proposes the Recombination byTransformation Genetic Algorithm (RTGA) that brings a newgenetic operator based on recombination by transformation ofbacteria. The same was compared with other variants of the AGin the optimization of four evaluation functions and it obtainedall the best overall optimum.

Ransomware is a subset of malware that is growing as a serious cyber threat. This malicious software prevents orlimits users from accessing their system until the ransom is paid.The use of Machine Learning (ML) algorithms has been widely used in automatic classification of these attacks. In this paper,we apply the Principal Component Analysis (PCA) techniqueas feature extraction intending to reduce dimensionality of the dataset, then we explore 11 ML algorithms in order to findthe best classifier for ransomware detection. Five comparisonmethods used in the literature were discussed. Nayes Bayesmethod achieved an Accuracy of 100% in one of the methods.