# Package gov.sandia.cognition.learning.function.scalar

Provides functions that output real numbers.

See: Description

• Interface Summary
Interface Description
PolynomialFunction.ClosedForm
Describes functionality of a closed-form algebraic polynomial function
• Class Summary
Class Description
AtanFunction
Returns the element-wise arctangent of the input vector, compressed between -maxMagnitude and maxMagnitude (instead of just -PI/2 and PI/2)
CosineFunction
A closed-form cosine function.
HardSigmoidFunction
A hard sigmoid function, which is an approximation of a logistic sigmoid whose output is between 0 and 1.
HardTanHFunction
A hard sigmoid function, which is an approximation of a tanh sigmoid whose output is between -1 and 1.
IdentityScalarFunction
A univariate scalar identity function: f(x) = x.
KernelScalarFunction<InputType>
The `KernelScalarFunction` class implements a scalar function that uses a kernel to compute its output value.
KolmogorovSmirnovEvaluator
You can specify a particular CDF.
LeakyRectifiedLinearFunction
A leaky rectified linear unit.
LinearCombinationScalarFunction<InputType>
A weighted linear combination of scalar functions.
LinearDiscriminant
LinearDiscriminant takes the dot product between the weight Vector and the input Vector.
LinearDiscriminantWithBias
A LinearDiscriminant with an additional bias term that gets added to the output of the dot product.
LinearFunction
This function acts as a simple linear function of the form f(x) = m*x + b.
LinearVectorScalarFunction
The `LinearVectorScalarFunction` class implements a scalar function that is implemented by a linear function.
LocallyWeightedKernelScalarFunction<InputType>
The `LocallyWeightedKernelScalarFunction` class implements a scalar function that uses kernels and does local weighting on them to get the result value.
PolynomialFunction
A single polynomial term specified by a real-valued exponent.
PolynomialFunction.Cubic
Algebraic treatment for a polynomial of the form y(x) = q0 + q1*x + q2*x^2 + q3*x^3
PolynomialFunction.Linear
Utilities for algebraic treatment of a linear polynomial of the form y(x) = q0 + q1*x
Utilities for algebraic treatment of a quadratic polynomial of the form y(x) = q0 + q1*x + q2*x^2.
PolynomialFunction.Regression
Performs Linear Regression using an arbitrary set of PolynomialFunction basis functions
RectifiedLinearFunction
A rectified linear unit, which is the maximum of its input or 0.
SigmoidFunction
An implementation of a sigmoid squashing function.
SoftPlusFunction
A smoothed approximation for rectified linear unit.
TanHFunction
The hyperbolic tangent (tanh) function.
ThresholdFunction
Maps the input space onto the set {LOW_VALUE,HIGH_VALUE}.
VectorEntryFunction
An evaluator that returns the value of an input vector at a specified index.
VectorFunctionLinearDiscriminant<InputType>
This class takes a function that maps a generic InputType to a Vector.
VectorFunctionToScalarFunction<InputType>
The `VectorFunctionToScalarFunction` class implements an adapter for using a vector function that outputs a single-dimensional vector as a scalar function.
VectorFunctionToScalarFunction.Learner<InputType>
The `VectorFunctionToScalarFunction.Learner` class implements a simple learner for a `VectorFunctionToScalarFunction` that allows a learning algorithm that outputs a vector function to be adapted to learn on data whose output are doubles.

## Package gov.sandia.cognition.learning.function.scalar Description

Provides functions that output real numbers. Scalar functions map some value onto the real numbers. This includes both functions that are the output of regression and functions such as squashing functions that are used within neural networks.
Since:
2.0
Author:
Justin Basilico