Please enable JavaScript.
Coggle requires JavaScript to display documents.
Provost Chapter 14 (Fundamental Concepts of Data Science (Data science…
Provost Chapter 14
Fundamental Concepts of Data Science
Data science team must keep in mind the problem that is being solved
Data should be considered an Asset, be careful on the investments made
Expected Value Framework can help us structure business problems
Be Careful of Over fitting and generalizations
Applying data science to a structured problem is different than exploratory analysis
Concepts of knowledge extraction from data
Identifying informative attributes, that have correlations
Fitting a numeric function model by choosing a target objective
Calculating similarities between objects in Data
Humans and Computers are good at different Things (Data science involves combining them)
Humans are better at identifying from out in the world variables (small relevant sets)
Only humans can tell best objective criterion and model in the end
computers are better at sifting through large quantities of data (Data robot)
Understand how data science fits into an organizations landscape
Thinking of Data science as concepts allows it to be applied to any general business problem
Data Scientists must keep up with the changing of technology to efficiently mine data
Our notion of the problem will often change the deeper we explore concrete problems (remember be creative)
explaining this change is crucial to keeping investors and CEO's on board
ETHICS!!!
always be forthcoming with your customers on your data collection
Target Leakage(Don't Forget It)
we must always ask ourselves is there enough data to solve the problem