Next:
Why Machine Learning
Typical Datamining Task
Datamining Result
Credit Risk Analysis
Other Prediction Problems
Problems Too Difficult to Program by Hand
Software that Customizes to User
Where Is this Headed?
What is the Learning Problem?
Type of Training Experience
Choose the Target Function
Possible Definition for Target Function
V
Choose Representation for Target Function
A Representation for Learned Function
Obtaining Training Examples
Learning Mechanism: Function Approximation
Final Design
Design Choices
Some Issues
Training Examples for
EnjoySport
Prototypical Concept Learning Task
What Justifies this Inductive Leap?
An UNBiased Learner
Inductive Bias
Inductive Systems and Equivalent Deductive Systems
