“An analysis of vulnerable IoT devices connected to the internet”

I wish to analyse the vulnerabilities which are found on smart devices found through machine learning techniques. Description of project to follow: Specifically beginning with a fingerprinting method to gather publically available IoT devices that are online. Identify and verify they are IoT – whether it is based on protocols, ports or a combination of parameters for identification. (product manuals for devices will give details on what is of importance for each) - From dataset, using a signature and or classification based machine learning algorithm to do the following. § Create classifier which works with a dataset, analysis of this dataset giving results of if IoT devices are secure or not. § Classify a range of security levels – A scale of security included which analyses the devices found and from that, giving results from 0-5 Vulnerabilities (0 – secure, 1 – moderately secure, 2 – Substantial, 3 – Severely insecure, 4 – Critically insecure) § Results include the above 0-4, and, the insecurity found ( ie. It uses ‘xyz’ protocol which is insecure leading it into the ‘xyz’ category. (Category meaning security/vulnerability level of 0-4)