Reading...
Play button
Play button
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;
6 Cards in this Set
- Front
- Back
|
What are the advantages of distributed memory?
|
1) resistance to damage. Redundancy means that death of individual cells have little effect.
2) Economy - single memories are stored in many cells, and each cell can participate in many memories. |
|
|
What are the 4 main simplifications of the McCulloch-Pitts neuron?
|
1) neuronal firing is a discrete all or nothing event
2) MCP works in discrete time 3) inhibition is all or nothing 4) the structure of the network is fixed |
|
|
What are the 5 main features of which an MCP neuron comprises
|
1) a set of excitatory inputs
2) a set of inhibitory inputs 3) a threshold value 4) a unit step activation function 5) a single output |
|
|
network models since Rosenblatt and hebb can be classified into basic kinds, give those kinds.
|
3 kinds
1) layered 2) lattice 3) recurrent |
|
|
what are the main features of a layered network
|
There are 3 main features of a layered network, they are;
1) no links between units in a layer 2) every unit in some layer is connected to every unit in the next layer 3) one-way flow of activiation |
|
|
What are the advantages and disadvantages of small and large values for the learning constant ?
|
large values allow for rapid training, however the lack of granularity means weights can oscillate between values jumping over the better value.
small values allow much less change of overshoot when training. however training will be slower as more iterations are required. |
