nuget
This class contains the various parameters which can affect the way in which an SVM is learned. Unless you know what you are doing, chances are you are best off using the default values.
Parameter
01 Syntax
Microsoft.VisualBasic.MachineLearning.SVM.Parameter
02 Methods
| Name | Overloads | Summary |
|---|---|---|
| .ctor | 1 | Default Constructor. Gives good default values to all parameters. |
| ToString | 1 | Display this parameter set as a json string. |
| Equals | 1 | Compares this parameter set with another object. |
| GetHashCode | 1 | Gets the hash code of this parameter set, which is combined by the hash code of all of its parameter values. |
| Clone | 1 | Creates a memberwise clone of this parameters object. |
03 Properties
| Name | Overloads | Summary |
|---|---|---|
| svmType | 1 | Type of SVM (default C-SVC) |
| kernelType | 1 | Type of kernel function (default Polynomial) |
| degree | 1 | Degree in kernel function (default 3). |
| gamma | 1 | Gamma in kernel function (default 1/k) |
| coefficient0 | 1 | Zeroeth coefficient in kernel function (default 0) |
| cacheSize | 1 | Cache memory size in MB (default 100) |
| EPS | 1 | Tolerance of termination criterion (default 0.001) |
| c | 1 | The parameter C of C-SVC, epsilon-SVR, and nu-SVR (default 1) |
| weights | 1 | ColorClass.name |
| nu | 1 | The parameter nu of nu-SVC, one-class SVM, and nu-SVR (default 0.5) |
| P | 1 | The epsilon in loss function of epsilon-SVR (default 0.1) |
| shrinking | 1 | Whether to use the shrinking heuristics, (default True) |
| probability | 1 | Whether to train an SVC or SVR model for probability estimates, (default False) |
04 Fields
| Name | Overloads | Summary |
|---|---|---|
| m_Weights | 1 | Contains custom weights for class labels. |
05 Members
#ctor
Default Constructor. Gives good default values to all parameters.
ToString
Display this parameter set as a json string.
Returns
A json text which describes all of the parameter values.
Equals(
Object)Compares this parameter set with another object.
Parameters
| Name | Type | Description |
|---|---|---|
obj | Object | The object that will be compared with this parameter set. |
Returns
True when the obj is a Parameter object which has the identical parameter values, otherwise False.
GetHashCode
Gets the hash code of this parameter set, which is combined by the hash code of all of its parameter values.
Returns
An Int32 hash code value.
Clone
Creates a memberwise clone of this parameters object.
Returns
The clone (as type Parameter)
svmType
Type of SVM (default C-SVC)
kernelType
Type of kernel function (default Polynomial)
degree
Degree in kernel function (default 3).
gamma
Gamma in kernel function (default 1/k)
Remarks
这个参数比较重要，千万不可以设置为零，否则将无法进行数据分类
coefficient0
Zeroeth coefficient in kernel function (default 0)
cacheSize
Cache memory size in MB (default 100)
EPS
Tolerance of termination criterion (default 0.001)
c
The parameter C of C-SVC, epsilon-SVR, and nu-SVR (default 1)
weights
ColorClass.name
Returns
A dictionary which maps the class name to its weight value, so that different penalty can be applied on different classes.
nu
The parameter nu of nu-SVC, one-class SVM, and nu-SVR (default 0.5)
P
The epsilon in loss function of epsilon-SVR (default 0.1)
shrinking
Whether to use the shrinking heuristics, (default True)
probability
Whether to train an SVC or SVR model for probability estimates, (default False)
m_Weights
Contains custom weights for class labels. Default weight value is 1.