LEARNING FROM DATA
Lecture Ten
(Chapter 10, Notes;
Chapter 11, Textbook)
1
Chapter 10: Learning From Data
Outline





The Concept of Learning
Data Visualization
Neural Networks
The Basics
Supervised and Unsupervised Learning
Business Applications
Association Rules
Implications for Knowledge Management
2
Chapter 10: Learning From Data
The “Context” of Learning


The “value added” collaborative intelligence
layer of KM architecture
Relevant technologies are:
 Artificial Intelligence
 Experts Systems
 Case-Based Reasoning
 Data Warehousing
 Intelligent agents
 Neural Networks
3
Chapter 10: Learning From Data
The “Process” of Learning


A process of filtering and
transforming data into valid
and useful knowledge.
Automate via technology tools:
 provide a collaborative
learning environment for
participants
 enhance their ability to
understand the processes /
tasks they are dealing with
4
Chapter 10: Learning From Data
The “Goals” of Learning


Final goal is to improve the
qualities of communication and
decision making
Ways to achieve these goals:
 Verify hypotheses (formed
from accumulated knowledge)
 Discover new patterns in data
 Predict future trends and
behaviour
5
Chapter 10: Learning From Data
Learning from Data



Build learning models that automatically
improve with experience.
Top-down approach
 Generate ideas
 Develop models
 Validate models
Bottom-up approach
 Discover new (unknown) patterns
 Find key relationships in data
6
Chapter 10: Learning From Data
Top-down approach (Example)




Start with a hypothesis derived
from observation or prior
knowledge
“Tourists visiting Egypt earn an
annual income of at least
$50,000”
Hypothesis tested by querying
database followed by analysis
If tests not supportive,
hypothesis is revised and test
7
Chapter 10: Learning From Data
Bottom-up approach (Example)




No hypothesis to test
“Find unknown buying
patterns by analyzing the
shopping basket”
“ … showed married males,
age 21 to 27, shopped for
diapers also brought beer.
“store decided to stack beer
cases next to diaper shelf”
8
Chapter 10: Learning From Data
Data Visualization


Explore visually for trends in
data useful for making decision
Exploring data includes:
 Identify key attributes and
their distribution
 Identify outliers
 Extract interesting grouping
of data subsets
 Identify initial hypothesis
9
Chapter 10: Learning From Data
Example of Data Visualization
(John Snow and the Cholera outbreak in London, 1845)
10
Chapter 10: Learning From Data
Artificial Neural Network as
Learning Model



Modeled after human brain’s
network
Simulate biological
information processing via
networks of interconnected
neurons
Neural networks are analog
and parallel
11
Chapter 10: Learning From Data
Neurons – The Basic Elements


The neuron receives
inputs, determines their
weights (strengths), sums
the combined inputs, and
compares the total to a
threshold (transfer
function)
If total is greater than
threshold, the neuron fires
(sends an output)
12
Chapter 10: Learning From Data
A Neuron Model
13
Chapter 10: Learning From Data
Learning in Neural Network


Supervised
 The NN needs a teacher
with a training set of
examples of input and
output
Unsupervised (or SelfSupervised)
 Does not need a teacher
14
Chapter 10: Learning From Data
Supervised Learning

Each element in a training
set is paired with an
acceptable response

Network makes successive
passes through the
examples

The weights adjust toward
the goal state.
15
Chapter 10: Learning From Data
A Supervised Neural Network
(An Example)
16
Chapter 10: Learning From Data
Unsupervised Learning

No external factors can
influence adjustment of
input’s weights

No advanced indication
of correct or incorrect
answers

Adjusts through direct
confrontation with new
experiences
17
Chapter 10: Learning From Data
Business Applications (1)

Risk management
 Appraise
commercial loan applications
 NN
trained on thousands of applications,
half of which were approved and the other
half rejected by the bank’s loan officers
 Through
supervised learning, NN learned
to pick risks that constitute a bad loan
 Identifies
loan applicants who are likely to
default on their payments
18
Chapter 10: Learning From Data
Business Applications (2)

Predicting Foreign Exchange Fluctuations:
 A set of relevant indicators were identified,
used as inputs to NN
 NN was trained for exchange rates of US
dollar against Swiss franc and Japanese yen,
using data from first 6 months of 1990.
Then it was tested over an 8- to 11-week
period
 Results revealed return on capital of about
20%
19
Chapter 10: Learning From Data
Business Applications (3)

Mortgage Appraisals:
 Neural
network uses the data in the
mortgage loan application
 It
estimates value of the property based
on the immediate neighborhood, the
city, and the country
 The
system comes up with a valuation
for the property and a risk analysis for
the loan.
20
Chapter 10: Learning From Data
Association Rules

A KB tool that generates a set of
rules to help understanding
relationships that exist in data

Types:
 Boolean
rule
 Quantitative
rule
 Multi-dimensional
 Multi-level
rule
association rule
21
Chapter 10: Learning From Data
Boolean Rule (An Example)

A rule that examines the
presence or absence of items

For example, if a customer
buys a PC and a 17” monitor,
then he will buy a printer.
Presence of items (a PC and
17” monitor) implies
presence of the printer in
the customer’s buying list
22
Chapter 10: Learning From Data
Quantitative Rule (An
Example)

A rule that considers the
quantitative values of items

For example, if a customer
earns between $30,000 and
$50,000 and owns an
apartment worth between
$250,000 and $500,000, he will
buy a 4-door automobile
23
Chapter 10: Learning From Data
Multi-dimensional Rule



A rule that refers to a
multitude of dimensions
If a customer lives in a big
city and earns more than
$35,000, then he will buy a
cellular phone
This rule involves 3
attributes: living, earning,
and buying. Therefore, it is
a multi-dimensional rule
24
Chapter 10: Learning From Data
Multi-level Association Rule
25
Chapter 10: Learning From Data
Implications for Knowledge
Management


Cost / benefit analysis
 Tangible costs - user training,
hardware + software, backup,
support, maintenance
 Intangible costs - user resistance and
learning curve
Quality Assurance
 Adequacy of initial design
 Level and frequency of maintenance
26
Chapter 10: Learning From Data
.....
Layers of KM Architecture
1
2
3
User Interface
(Web browser software installed on each user’s PC)
Authorized access control
(e.g., security, passwords, firewalls, authentication)
Collaborative intelligence and filtering
(intelligent agents, network mining, customization, personalization)
Knowledge-enabling applications
4
5
(customized applications, skills directories, videoconferencing, decision support systems,
group decision support systems tools)
Transport
(e-mail, Internet/Web site, TCP/IP protocol to manage traffic flow)
Middleware
6
(specialized software for network management, security, etc.)
The Physical Layer
(repositories, cables)
7
Databases
Legacy applications
(e.g., payroll)
Groupware
(document exchange,
collaboration)
Data warehousing
(data cleansing,
data mining)
27