The NeuralNet_2L companion object provides factory methods for creating two-layer (no hidden layer) neural networks. Note, 'scale' is defined in Scaling.
Attributes
- Companion
- class
- Graph
-
- Supertypes
- Self type
-
NeuralNet_2L.type
Members list
Value members
Concrete methods
Create a NeuralNet_2L with automatic rescaling from a combined data matrix.
Create a NeuralNet_2L with automatic rescaling from a combined data matrix.
Value parameters
- col
-
the first designated response column (defaults to the last column)
- f
-
the activation function family for layers 1->2 (input to output)
- fname
-
the feature/variable names (defaults to null)
- hparam
-
the hyper-parameters (defaults to Optimizer.hp)
- xy
-
the combined input and output matrix
Attributes
Create a NeuralNet_2L with automatic rescaling from a data matrix and response vector. As the number of output nodes is one in this case, it is effectively a perceptron.
Create a NeuralNet_2L with automatic rescaling from a data matrix and response vector. As the number of output nodes is one in this case, it is effectively a perceptron.
Value parameters
- f
-
the activation function family for layers 1->2 (input to output)
- fname
-
the feature/variable names (defaults to null)
- hparam
-
the hyper-parameters (defaults to Optimizer.hp)
- x
-
the input/data m-by-n matrix
- y_
-
the output/response m-vector
Attributes
Create a NeuralNet_2L with automatic rescaling from a data matrix and response matrix.
Create a NeuralNet_2L with automatic rescaling from a data matrix and response matrix.
Value parameters
- f
-
the activation function family for layers 1->2 (input to output)
- fname
-
the feature/variable names (defaults to null)
- hparam
-
the hyper-parameters (defaults to Optimizer.hp)
- x
-
the input/data m-by-n matrix
- y
-
the output/response m-by-ny matrix