MADMAX: Browser-Based Malicious Domain Detection Through Extreme Learning Machine

Fast and accurate malicious domain detection is an essential research theme to prevent cybercrime, machine learning attractive approach for detecting unseen domains in the past decade. In this paper, we present MADMAX (MAchine learning-baseD MAlicious eXhauster) , a browser-based application leveraging extreme (ELM) detection. contrast existing work of ELM-based detection, newly introduces two methods, i.e., selection optimized features provide higher accuracy throughput based on permutation importance real-time training retrain model with updated dataset continuous We demonstrate that fairly outperforms respect by virtue features. Moreover, also confirm stably detects even domains, whereas without decreases due domains. The source codes publicly available via GitHub.