Desktop Survival Guide
by Graham Williams


Temporal Difference Learning: Learning based on the difference between temporally successive predictions, rather than error between predicted and actual values, so that the learner's current prediction for the current input data more closely matches the next prediction at the next time step, thus learning occurring as predictions change over time. Forms the basis for Reinforcement Learning.

Terminological Logics: Formalise the notion of Frames as structured types, often called Concepts. These logics include a set of syntactic constructs that form concepts, and other, related, notions such as roles, and are based on formal model-theoretic semantics which provide firm definitions for the syntactic constructs of the logic. Synonyms include Description Logic and Concept Language.

Test Set: A portion of a dataset used to test the performance of a model. See also Training Set and Validation Set.

Training Set: A portion of a dataset that is used to build a model. See also Test Set and Validation Set.

Copyright © 2004-2006
Support further development through the purchase of the PDF version of the book.
Brought to you by Togaware.