Initializer
The Initializer object provides functions to initialize the parameters/weights of Neural Networks. Supports Uniform, Normal and Nguyen & Widrow methods.
Attributes
- Graph
-
- Supertypes
-
class Objecttrait Matchableclass Any
- Self type
-
Initializer.type
Members list
Value members
Concrete methods
Generate a random weight/parameter matrix with elements values in (0, limit).
Generate a random weight/parameter matrix with elements values in (0, limit).
Value parameters
- cols
-
the number of columns
- limit
-
the maximum value for any weight
- rows
-
the number of rows
- stream
-
the random number stream to use
Attributes
Generate a random weight/parameter matrix with elements values from the Standard Normal distribution.
Generate a random weight/parameter matrix with elements values from the Standard Normal distribution.
Value parameters
- cols
-
the number of columns
- rows
-
the number of rows
- stream
-
the random number stream to use
Attributes
Generate a random weight/parameter matrix with elements values from the Nguyen & Widrow method.
Generate a random weight/parameter matrix with elements values from the Nguyen & Widrow method.
Value parameters
- cols
-
the number of columns
- rows
-
the number of rows
- stream
-
the random number stream to use
Attributes
- See also
-
ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=6945481
Generate a random weight/parameter vector with elements values in (0, limit).
Generate a random weight/parameter vector with elements values in (0, limit).
Value parameters
- limit
-
the maximum value for any weight
- rows
-
the number of rows
- stream
-
the random number stream to use
Attributes
Generate a random weight/parameter matrix with elements values from the Standard Normal distribution.
Generate a random weight/parameter matrix with elements values from the Standard Normal distribution.
Value parameters
- rows
-
the number of rows
- stream
-
the random number stream to use
Attributes
Generate a random weight/parameter matrix with elements values from the Nguyen & Widrow Method.
Generate a random weight/parameter matrix with elements values from the Nguyen & Widrow Method.
Value parameters
- rows
-
the number of rows
- stream
-
the random number stream to use
Attributes
- See also
-
ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=6945481