Artificial IntelligenceMcGraw-Hill, 1991 - 621 páginas |
Dentro del libro
Resultados 1-3 de 32
Página 504
... output layer . During the backward pass , the network's actual output ( from the forward pass ) is compared with the target output and error estimates are computed for the output units . The weights connected to the output units can be ...
... output layer . During the backward pass , the network's actual output ( from the forward pass ) is compared with the target output and error estimates are computed for the output units . The weights connected to the output units can be ...
Página 505
... output vector is y . Assign activation levels to the input units . 5. Propagate the activations from the units in ... output layer . 1 B 1 + e Σ2 , for all j = 1 , ... , C Again , the thresholding weight w2oj for output unit j plays a ...
... output vector is y . Assign activation levels to the input units . 5. Propagate the activations from the units in ... output layer . 1 B 1 + e Σ2 , for all j = 1 , ... , C Again , the thresholding weight w2oj for output unit j plays a ...
Página 512
... output units becomes overwhelming . Soon , the other output units are all completely inactive . This type of mutual inhibition is called winner - take - all behavior . One popular unsupervised learning scheme based on this behavior is ...
... output units becomes overwhelming . Soon , the other output units are all completely inactive . This type of mutual inhibition is called winner - take - all behavior . One popular unsupervised learning scheme based on this behavior is ...
Contenido
5 | 24 |
Heuristic Search Techniques | 63 |
Knowledge Representation Issues | 105 |
Derechos de autor | |
Otras 28 secciones no mostradas
Otras ediciones - Ver todas
Términos y frases comunes
Abbott agents algorithm answer apply approach ARMEMPTY assertions attributes axioms backpropagation backtracking backward belief best-first search breadth-first search Caesar called Chapter chess clauses complete concept conceptual dependency consider constraints contains contradiction corresponding define depth-first depth-first search described discussed domain fact frame function game tree goal grammar graph heuristic Horn clauses important inference inheritance input instance interpretation isa links John justification knowledge base knowledge representation labeled learning Marcus match minimax move MYCIN natural language node object ON(B operators output parsing particular path perceptron perform players possible preconditions predicate logic problem problem-solving procedure produce PROLOG properties represent result robot rules script Section semantic semantic net sentence shown in Figure simple slot solution solve specific step structure Suppose syntactic task techniques theorem things tree truth maintenance system understanding variables version space