Types of Data Mining Methods
Supervised Learning
(Predictive Analytics) Unsupervised Learning
• Prediction (numerical Y) • Segmentation/Clustering
• Classification (categorical Y) • Relationship Mining
• Recommender System
• …
Business Analytics using
Data Mining in a Nutshell
Prof. Vandith Pamuru
• How is Data Mining
similar to or different
from Traditional
Objectives in Statistics?
3
What can you do/say if you had a
data set on past house sales?
Price SqFt Bedrooms Bathrooms Offers Brick Neighborhood
114300 1790 2 2 2 No East
114200 2030 4 2 3 No East
114800 1740 3 2 1 No East
94700 1980 3 2 3 No East
119800 2130 3 3 3 No East
114600 1780 3 2 2 No North
151600 1830 3 3 3 Yes West
150700 2160 4 2 2 No West
Statistics Data Mining
• Macro decisions • Micro decisions
– On average what is – What is going to
going on in the happen to AN
whole individual
population/houses? entity/house?
• Estimate and interpret • Predict the sale price of
the pricing structure of a house that is on the
houses in the city market
5
Statistics Data Mining
• Macro decisioning • Micro decisioning
• Explain/describe • Predict values of new
population relationships records
• Small sample, few • Large sample, many
variables variables
What will you predict? In order
to decide what?
7
Decision vs. Prediction
• What will you predict? In order to
decide what?
– Identify an aspect that is not known
beforehand, however would have helped
in making a better decision if it was
known
– Can you predict the unknown?
8
How to Build and Evaluate a
Model?
What decisions
are involved to Does your
achieve the What do prediction
business you model optimize
objective? predict? the business
objective?
Business Data Mining Model
Objective Problem Evaluation
Alignment
In the following examples,
• What is the business objective?
• What is the business decision?
• What is the data mining problem?
What will you do if you had a data on
performance of house loans ─ defaulted
vs. paid in full?
What will you predict? In order to decide what?
Scope for predictive analytics is
enormous
Defeating Crimes and Cybercrimes
Risk and Fraud Analytics
