Training a Single Sigmoidal Neuron Is Hard (Q4409384): Difference between revisions

@@ Property / cites work @@
+The hardness of approximate optima in lattices, codes, and systems of linear equations
+Normal rank
@@ Property / cites work @@
+Learnability and the Vapnik-Chervonenkis dimension
+Normal rank
@@ Property / cites work @@
+On the geometric separability of Boolean functions
+Normal rank
@@ Property / cites work @@
+Computational limitations on training sigmoid neural networks
+Normal rank
@@ Property / cites work @@
+Robust trainability of single neurons
@@ Property / cites work: Robust trainability of single neurons / rank @@
+Normal rank
@@ Property / cites work @@
+The densest hemisphere problem
@@ Property / cites work: The densest hemisphere problem / rank @@
+Normal rank
@@ Property / cites work @@
+The computational intractability of training sigmoidal neural networks
+Normal rank
@@ Property / cites work @@
+On the complexity of loading shallow neural networks
+Normal rank
@@ Property / cites work @@
+On the complexity of polyhedral separability
@@ Property / cites work: On the complexity of polyhedral separability / rank @@
+Normal rank
@@ Property / cites work @@
+Computational limitations on learning from examples
+Normal rank
@@ Property / cites work @@
+Learning representations by back-propagating errors
+Normal rank
@@ Property / cites work @@
+Feedforward nets for interpolation and classification
+Normal rank