Intelligent Systems

1
Intelligent Systems
2
A Diverse Field

• theoretical
• What is intelligence?
• ambitious
• Lets build an artificial (enhanced) human
  being.
• practical
• How can we discover a set of good rules that
  predict which customers will repay their loans?

3
Three Hot Areas of Research

• autonomous vehicles
• semantic technologies
• data mining (machine learning)

4

• Autonomous Vehicles

5
Autonomous Vehicles

• DARPA Grand Challenge
• autonomous vehicles traverse a desert route
• http//www.darpa.mil/grandchallenge/index.html

6
2004 Results

• greatest distance traveled was 7.36 miles
• 1 million prize not awarded

7
2005 Course

• http//www.darpa.mil/grandchallenge05/gcorg/index.
  html

8
2005 Results

• five vehicles completed a 131.2-mile course, four
  of them in under ten hours
• Stanford Universitys Stanley finished in 6
  hours, 53 minutes to win the 2 million prize

9
The Winner

• Photo from DARPA

10

• Semantic Technologies

11
The Data Glut

• we are generating massive amounts of data
• want computers to process data in more
  sophisticated ways
• need techniques that enable computers to attach
  meaning to information

12
Semantic Technologies

• design semantic models (ontologies) for
  application domains
• provide explicit meaning for information
• basic concepts and their relationships
• use logic and inference to reason about the
  domain knowledge
• these techniques extend
• databases
• object models
• business rules
• XML schemas

13
Semantic Technologies

• 2 billion industry today
• expected to grow very rapidly

14
The Semantic Web

• World Wide Web Consortium standards
• RDF (resource description framework)
• OWL (the web ontology language)

15
Example

• ltowlClass rdfID"Pasta"gt
• ltrdfssubClassOf rdfresource"EdibleThing"
  /gt
• ltowldisjointWith rdfresource"Meat" /gt
• ltowldisjointWith rdfresource"Fowl" /gt
• ltowldisjointWith rdfresource"Seafood" /gt
• ltowldisjointWith rdfresource"Dessert" /gt
• ltowldisjointWith rdfresource"Fruit" /gt
• lt/owlClassgt

16

• Data Mining

17
Data Mining

• machine learns a model for a data set
• discovers patterns in the data
• model can be used to
• help understand the data
• make predictions about future data
• will a customer repay a loan?

18
Example Models

• rules
• decision trees
• neural networks

19
Weka

• a popular (free) data mining tool developed at
  the University of Waikato in New Zealand
• http//www.cs.waikato.ac.nz/ml/weka/index.html

20
Weka

• incorporates all of the standard data mining
  models
• written in Java
• provides an API that can be accessed from Java
  code

21
Iris Classification

• http//vtgcrec.ifas.ufl.edu/pages/Florida20Botani
  cal_gardens-4-03.htm

22
The Data

• _at_RELATION iris
• _at_ATTRIBUTE sepallength REAL
• _at_ATTRIBUTE sepalwidth REAL
• _at_ATTRIBUTE petallength REAL
• _at_ATTRIBUTE petalwidth REAL
• _at_ATTRIBUTE class Iris-setosa,Iris-versicolor,Iri
  s-virginica
• _at_DATA
• 5.1,3.5,1.4,0.2,Iris-setosa
• 7.0,3.2,4.7,1.4,Iris-versicolor
• 6.3,3.3,6.0,2.5,Iris-virginica

file contains data for 150 specimens
23
The Data Mining Task

• find a model that predicts the class given the
  petal and sepal attributes

24
Demo

• use Weka to find a decision tree model for iris
  classification

25
Intelligent Systems

• The Beginning