Test 4 Deep Learning Flashcards
(8 cards)
1
Q
1964
A
Tom Hover - Theorem that if you randomly assign binary labels, the probability of convergence is 0 if N>P, ~1 otherwise
2
Q
1969
A
Seymour Report and Minsky- Criticizes perceptrons for not converging
3
Q
1970s
A
Name Change - Adaptive Filters
4
Q
1973
A
Lighthall Report - AI research not promising (UK Govt)
5
Q
1982
A
Hopfield - used “energy” of atoms and compared it to neuron, minimize the energy.
6
Q
1982
A
Hopfield Net - associative memory using Mcculloch/pitts
7
Q
1983
A
Hinton/Semonsky - Adaptive Stochastic Hopfield Nets with hidden units
8
Q
1986
A
Williams introduced backpropagation
