(eng): Modelling network traffic is gaining importance to counter modern security threats of ever increasing sophistication. It is though surprisingly difficult and costly to construct reli­able classifiers on top of telemetry data due to the variety and complexity of signals that no human can manage to interpret in full. Obtaining training data with sufficiently large and variable body of labels can thus be seen as a pro­hibitive problem. The goal of this work is to detect infected computers by observing their HTTP(S) traffic collected from network sensors, which are typically proxy servers or net­work firewalls, while relying on only minimal human input in the model training phase. We propose a discriminative model that makes decisions based on a computer’s all traf­.c observed during a predefined time window (5 minutes in our case).