The Hillstone Advanced Threat Detection (ATD) Engine analyzes millions of known HTTP based malware samples with regular updates. It extracts the common characteristics of each malware family for further analysis. Using unsupervised machine learning algorithms and mathematical modeling, the common features for each of the thousands of malware families is poured into a clustering model and is loaded on the device.
The model is also regularly updated from Hillstone Cloud . This process is illustrated in the following diagram