Using machine learning to identify botnets and non-bonet traffic

I need the development of a solution which is GUI based and easy to use that uses machine learning algorithms to identify botnets and non-botnet (benign) traffic. The first step in the project is to extract HTTP/HTTPS packets as flows from pcap files. The flow will have to be extracted from both botnet and non-botnet traffic. There are some software that can be used to extract the features. Once this is done, these will be loaded into a Pandas frame and labelled (as botnet and non-botnet) and then the features need to be extracted. The features will then be selected, and this will be done using a manual and automated method. I want to use static feature selection method and also several automated feature selection algorithms to identify the best features that can be used by the machine learning algorithms to identify botnet and non-botnet traffic. I want to be able to compare the result of using an automated feature selection algorithm and a manual feature selection algorithm to see if a manual feature selection method can compare to automated feature selection algorithms. This must also be tested on a new unseen dataset as well and also on some additional variants of the Botnet and the results compared.