Please enable JavaScript.
Coggle requires JavaScript to display documents.
Date Science for Business By: Fawcett & Provost "Chapter 14:…
Date Science for Business
By: Fawcett & Provost
"Chapter 14: Conclusion"
The Fundamental
Concepts of Data Science
Fitting in the organization
and competitive landscape
Attract, structure, & nurture
data science teams
Data science leads
to competitive advantage
Tactical principles
for doing well
Thinking
data-analytically
Data mining
process
Collection of different
high-level science tasks
Generalization
and overfitting
Expected value
framework
Extracting knowledge
from data
Identifying informative
attributes
Numeric function
models to data
Controlling complexity
Calculating similarities
Applying our Fundamental Concepts
to a New Problem: Mining Mobile Device Data
Going beyond
exploratory data
Visual representation
(i.e. scatterplot)
Conceptual toolkits
Helpful in thinking
about a brand-new problem
Data-understanding
phase
Decide how exactly
to represent the variables
Mining data on
documents
Treat each document
as a collection of words
Changing the Way We Think about
Solutions to Business Problems
Be conscious of changes
to fit the data
All stakeholders are not
involved with the data science
problem formulation
Hitting TPs outweight
possibly targeting FPs
What Data Can't Do: Humans
in the Loop, Revisited
Consider the limits of
data science and data-driven
decision-making
Meaning of data is colored
by our own beliefs
Task selection and specification
are areas where human interaction
is critical
i.e. selecting the right
data to mine
Human interaction critical
in evaluation stage of process
Humans can tell
best objective criterion to optimize
Interpretation Changes
Humans often change
their understanding of "what the
data is"
What's the overall value?
We must consider data
as an asset
Privacy, Ethics, and Mining
Data About Individuals
Ethical issues should
not be ignored
A variety of stakeholders
are involved in data mining
Tensions between privacy and
improving business decisions
Direct relationship between increased
use of personal data and increased effectiveness
of the associated business decisions
Privacy concerns
Not easy-to-deal
with issues
Not easy-to-understand
Final Words
Bringing interactions with
stakeholders to life
Thinking carefully about
what is important to the
business problem
Considering evaluation and "baselines
for comparison
Data scientists: think deeply about why your
work is relevant to helping business
Business stakeholder: don't be confused
with data scientist jargon