RECH class
Create a RECH model object
Syntax
Mdl = RECH(Specs,Name,Value)
Description
Mdl = RECH(Specs,Name,Value)
returns RECH model object Mdl
given particular specification Specs
. Name
and Value
specify additional options using one or more name-value pair arguments. For example, users can specify different choices of priors for model parameters.
See: Input Arguments, Output Argument, Examples
Input Arguments
Data type: string
Specification of the RECH model, specified as one of these values:
'SRN-GARCH' | SRN-GARCH specification |
'SRN-GJR' | SRN-GJR specification |
'SRN-EGARCH' | SRN-EGARCH specification |
See Nguyen et al. (2020) for mathematical description of specification of the RECH model.
Name-Value Pair Arguments
Specify optional comma-separated pairs of Name,Value
arguments. Name
is the argument name and Value
is the corresponding value. Name
must appear inside quotes. You can specify several name and value pair arguments in any order as Name1,Value1,...,NameN,ValueN
.
Example: 'Prior',{{'alpha','beta'},'Uniform',[0,1]},'Distribution','Normal'
specifies that the prior distribution of parameters $\alpha$ and $\beta$ is a uniform distribution in $[0,1]$, and the innovation is standard normal.
'Prior' - Prior distribution of model parameters
Data Type: Cell Array
Prior distribution of the model parameters, specified as the comma-separated pair consisting of 'Prior' and a 2D cell array whose each row consists of:
- Parameter names, specified as a cell array of strings.
- Distribution name, specified as a string.
- Distribution parameters, specified as an array of numerical number.
The RECH class use VarName
property to store parameter names. The parameter names for each specification of the RECH model is defined as:
Specification | Parameter names |
---|---|
'SRN-GARCH' | 'alpha','beta','beta0','beta1','v0','v1','v2','w','b' |
'SRN-GJR' | 'alpha','beta','gamma','beta0','beta1','v0','v1','v2','w','b' |
'SRN-EGARCH' | 'alpha','beta','gamma','omega','beta0','beta1','v0','v1','v2','w','b' |
The VBLab package provides following options for prior distribution of the RECH model parameters
Prior distribution | Description |
---|---|
'Normal' | Normal distribution $\mathcal{N}(\mu,\sigma^2)$ |
'Gamma' | Gamma distribution $\Gamma(\alpha,\beta)$ |
'Inverse-Gamma' | Inverse-Gamma distribution $\Gamma^{-1}(\alpha,\beta)$ |
'Beta' | Beta distribution $\text{Beta}(\alpha,\beta)$ |
'Uniform' | Uniform distribution $\mathcal{U}(\alpha,\beta)$ |
Default: The default priors for each specification of the RECH model are as following
Specification | Prior |
---|---|
'SRN-GARCH' | {{'alpha','beta'},'Uniform',[0,1]; {'beta0','beta1'},'Normal',[0,1];{'v0','v1','v2','w','b'},'Normal',[0,0.1]} |
'SRN-GJR' | {{'alpha','beta'},'Uniform',[0,1]; {'beta0','beta1'},'Normal',[0,1];{'v0','v1','v2','w','b'},'Normal',[0,0.1]} |
'SRN-EGARCH' | {{'alpha','beta'},'Uniform',[0,1]; {'beta0','beta1'},'Normal',[0,1];{'v0','v1','v2','w','b'},'Normal',[0,0.1]} |
Example: 'Prior',{{'beta0','beta1'},'Normal',[0,0.1]}
Output Arguments
Data type: RECH Object
RECH is an object of the RECH class with pre-defined properties and functions.
Object Properties
The RECH object properties include information about model-specific information, coefficient estimates and fitting method.
Properties | Data type | Description{: .text-center} |
---|---|---|
ModelName | string (r) | Name of the model, which is 'RECH' |
VarName | cell array (r) | Model parameter names. Stored in a cell array of strings |
NumParams | integer (+) | Number of model parameters |
Post * | struct | Information about the fittedd method used to estimate model paramters • RECH models can only be fitted by MGVB technique |
Coefficients * | Cell array | • Estimated Mean of weights of Deep Neuron Network • Used to doing point estimation for new test data |
CoefficientVar * | cell array (r) | Variance of coefficient estimates |
LogLikelihood * | double (r) | Loglikelihood of the fitted model. |
Notation:
- * $\rightarrow$ object properties which are only available when the model is fitted. Default value is
None
. - (+) $\rightarrow$ positive number.
- (r) $\rightarrow$ read-only properties.
Object Methods
Use the object methods to initialize model parameters, predict responses, and to visualize the prediction.
vbayesInit | Initialization method of model parameters |
vbayesPredict | Predict responses of fitted DeepGLM models |
Examples
To be updated…
Reference
[1] Nguyen, T.-N., Tran, M.-N., and Kohn, R. (2020). Recurrent conditional heteroskedasticity. arXiv:2010.13061. Read the paper
See Also
LogisticRegression $\mid$ DeepGLM $\mid$ Custom model $\mid$ MGVB