With a data set established, we need analytical models to gain predictive insights. By looking at historical weaponization trends, we can train algorithms to look across diverse types of data and identify the combination of traits that best predicts which vulnerabilities will be weaponized by attackers in the wild. Just as importantly, this approach can predict the speed at which a given vulnerability is likely to be weaponized.

•  Traits and vulnerabilities - Couldn't we start with the threats that actually succeeded. Then take that information and categorize it using the risk score from CVE or others. take that information and create a relationship database (i.e SharedDB or No-SQL columnar DB) where big data comes into play to establish or identify those relationships, this will help the end-user determine the number of similarities between the variants or possible vulnerabilities that exist
• Locations - identify where the code is coming from by associating the geographic regions, with the code, actors and success levels, this allows for those models (again Big Data) to start narrowing down the attacks to specific regions based on the type of attack, its function, success rate and locale (determine the type of attack and method of attack based on their success rate and design).
• Finally, use ML to look at the attack vectors from a historical standpoint, the results from BigData can now inject its findings into the ML DB and from those relationshps, we can determine based on risk score if something else will occur as part of the variants evolution (most systems build on itself).