Please enable JavaScript.
Coggle requires JavaScript to display documents.
Conclusion (3 Fundamental concepts of data science (2) How to think…
Conclusion
3 Fundamental concepts of data science
1) How does DS fit the business
how will data science lead to a competitive advantage
helps sustain competitive advantage
2) How to think analytically
data mining process
collection of high level data science tasks
data is an asset, how do we leverage it
cost benefits of knowledge recieved
mitigate generalization and overfitting
apply data science to different tasks
3) Extracting knowledge from data
1) identify informative attributes
2) fitting a numeric function model to data
3) controlling complexity is important
4) calculating similarity between objects described by data
lift
how likely as pattern is opposed to chace
helps judge evidence for or against a conclusion
Applying fundamental concepts to a new problem
Mobile phone e commerce example
What data can not do
humans can sift through the large variables of the world to find those of importance
computers can take that data and examine it quickly
humans can evaluate end results and can tweak programs to optimize them
the data we are using is a result of humans
psychology is very important to understand
the data that is spit out might be confusing
seeking someone with domain expertise can be of great help
humans interpret the data very well
humans can implement solutions and make changes
data is great at getting us there
Final Words
having fundamental concepts is very important to give structure
when you have reached the edge it is then important to go back to the concepts
always be thinking about the problem to solve
know the different types of tasks needed
data scientist should be able to communicate to business professionals
Privacy, ethics, and data about individuals
there are ethical implications about data mining
privacy has many meanings
Privacy and improving business
direct conflict
less privacy
more improved data science
more privacy
less improved data science
current day data scientists need to work within means
Is there more to data science
there are many ways to go deeper into the data with advanced techniques
Final example
micro - outsorcing
outsourcing of large numbers of very small, well-defined tasks
keeping ads off of questionable site
this keeps public image of the company good
we can use data science for this task