6.034 Artificial Intelligence - Recitations, fall 2004 online slides on learning

Next: Prediction Problems Previous: The Learning Problem

Learning From Examples

*
Learn general concepts or categories from examples
*
Learn a task (drive a vehicle, win a game of backgammon)
*
Examples of objects or tasks are gathered and stored in a database
*
Each example is described by a set of attributes or features
*
Each example used for training is classified with its correct label (chair, not chair, horse, not horse, 1 vs. 2 vs. 3, good move, bad move, etc.)
*
The machine learning program learns general concept description from these specific examples
*
The ML program should be applied to classify or perform tasks never before seen from learned concept