Develop a module that, given the pre-processed dataset, returns a set of optimized models.

We need to create a module that loads the .pcap file(s) and extracts the contents of the Wi-Fi header, in particular:

• frame length
• MAC address
• Flags

This module should also extract the audio signal from the ground-truth recording.

This functionality should allow us to compare the network traffic with the audio data recordings.

We need to build a module that filters the packets that are Granger-caused by the ground-truth data series.

There should be a function that allows the project to read .wav audio files.

We need to build a module that filters the relevant packets. A more precise definition of "relevant" will follow.

Also, this module should do some basic filtering (like noise reduction) on the ground-truth audio recording.

Implement feature selection module

Some values are saved as NaN, for some obscure reason. CDF_pl seems to be the problem.

We must build a module to classify the MAC addresses into {Microphone, Not-Microphone}.
This module should test the following ML algorithms (list to be expanded):

• Decision Tree Classifier
• ExtraTrees
• RandomForest
• ExtraTrees
• SVM
• KNN