This is a Deep Neural Network based Host Intrusion Detection System, that can run on linux machines. It uses the CICIDS2017 dataset. And currently only supports detection of portscan attacks with an accuracy of ~98%. Has a GUI for basic user interaction and shows the neural network behavior as well as packet flows on the GUI. Has the ability to save the captured flow data as well as detected attack information in logs, for further processing. There's a lot more work that still needs to be done on this. This is a working proof-of-concept.