Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
11 Cards in this Set
- Front
- Back
|
Parallel Distributed Processing |
Type of connectionist modeling involving 8 major aspects: a set of processing units, an activation for each unit, an output for each unit, pattern of connectivity between units, a propagation rule spreading the activations via the connections, an activation rule for processing for combining inputs to a unit to determine its new activation, a learning rule for modifying activations based on experience, and an environment that provides the system with experience |
|
|
A set of processing units |
These would be nodes, demons, neurons, etc. They are represented by a set of integers |
|
|
Activation for each unit |
This would be like an action potential for the unit and it is represented by a vector of time dependent functions |
|
|
Output for each unit |
Each unit outputs to units it is connected to. This process is a function of activation |
|
|
Pattern of connectivity |
This is the strength of connections between units. In neuronal terms we could think number of AMPA receptors on post synaptic cell, in PDP models this is represented by a matrix of weighted numbers |
|
|
Activation Rule |
If a unit is activated it will output to all of the units its output center is connected to |
|
|
Learning Rule |
A system for modifying unit connection strength based on experience. Could be like when Glutamate stimulated creation of more AMPA receptors |
|
|
Environment |
Need something to provide the system with experience |
|
|
Viewpoint Invariance |
Part of Biederman's RBC theory saying we can recognize an object regardless of viewing angle due to its invariant edge properties |
|
|
Invariant Edge Properties |
Curvature (various points of a curve), Parallel lines (two or more points which follow the same direction), Co-termination (the point at which two points meet and therefore cease to continue), Symmetry and asymmetry Co-linearity (points branching from a common line) |
|
|
Biederman's study |
Showed images of objects missing invariant properties such as areas of co-termination which made the object difficult to identify |
