Accelerating Discovery

6 ◾ Accel erating Disc over y

way to put that powerful specialized processor to work on your data

problem and you will inevitably be astounded at what you can see.

• Everything is a taxonomy/feature vector/network: At the risk of

oversimplifying things a bit, there are really only three basic things

you need to know to make sense of data: What the entities are that

you care about and how they relate to each other (the taxonomy),

how you can describe those entities as features (feature vector), and

how you can represent the way those entities interact (network).

Every problem involves some subset or combination of these ideas. It

really is that simple.

• Time is not a magazine: e data we take to begin our investigation

with is usually ...