GMMN_trained {gnn}R Documentation

Trained Generative Moment Matching Networks

Description

Trained generative moment matching networks (GMMNs); see also the demos GMMN_QMC_paper and GMMN_MTS_paper.

Usage


data("GMMN_dim_2_300_2_ntrn_60000_nbat_5000_nepo_300_C_tau_0.25")
data("GMMN_dim_2_300_2_ntrn_60000_nbat_5000_nepo_300_C_tau_0.5")
data("GMMN_dim_2_300_2_ntrn_60000_nbat_5000_nepo_300_C_tau_0.75")
data("GMMN_dim_2_300_2_ntrn_60000_nbat_5000_nepo_300_G_tau_0.25")
data("GMMN_dim_2_300_2_ntrn_60000_nbat_5000_nepo_300_G_tau_0.5")
data("GMMN_dim_2_300_2_ntrn_60000_nbat_5000_nepo_300_G_tau_0.75")
data("GMMN_dim_2_300_2_ntrn_60000_nbat_5000_nepo_300_eqmix_C_tau_0.5_rot90_t4_tau_0.5")
data("GMMN_dim_2_300_2_ntrn_60000_nbat_5000_nepo_300_eqmix_G_tau_0.5_rot90_t4_tau_0.5")
data("GMMN_dim_2_300_2_ntrn_60000_nbat_5000_nepo_300_eqmix_MO_0.75_0.6_rot90_t4_tau_0.5")
data("GMMN_dim_2_300_2_ntrn_60000_nbat_5000_nepo_300_MO_0.75_0.6")
data("GMMN_dim_2_300_2_ntrn_60000_nbat_5000_nepo_300_t4_tau_0.25")
data("GMMN_dim_2_300_2_ntrn_60000_nbat_5000_nepo_300_t4_tau_0.5")
data("GMMN_dim_2_300_2_ntrn_60000_nbat_5000_nepo_300_t4_tau_0.75")
data("GMMN_dim_3_300_3_ntrn_60000_nbat_5000_nepo_300_NC21_tau_0.25_0.5")
data("GMMN_dim_3_300_3_ntrn_60000_nbat_5000_nepo_300_NG21_tau_0.25_0.5")
data("GMMN_dim_5_300_5_ntrn_60000_nbat_5000_nepo_300_C_tau_0.5")
data("GMMN_dim_5_300_5_ntrn_60000_nbat_5000_nepo_300_G_tau_0.5")
data("GMMN_dim_5_300_5_ntrn_60000_nbat_5000_nepo_300_NC23_tau_0.25_0.5_0.75")
data("GMMN_dim_5_300_5_ntrn_60000_nbat_5000_nepo_300_NG23_tau_0.25_0.5_0.75")
data("GMMN_dim_5_300_5_ntrn_60000_nbat_5000_nepo_300_t4_tau_0.5")
data("GMMN_dim_10_300_10_ntrn_60000_nbat_5000_nepo_300_C_tau_0.5")
data("GMMN_dim_10_300_10_ntrn_60000_nbat_5000_nepo_300_G_tau_0.5")
data("GMMN_dim_10_300_10_ntrn_60000_nbat_5000_nepo_300_NC55_tau_0.25_0.5_0.75")
data("GMMN_dim_10_300_10_ntrn_60000_nbat_5000_nepo_300_NG55_tau_0.25_0.5_0.75")
data("GMMN_dim_10_300_10_ntrn_60000_nbat_5000_nepo_300_t4_tau_0.5")
data("GMMN_dim_3_100_3_ntrn_4996_nbat_4996_nepo_1000_PCA_ZCB_USD")
data("GMMN_dim_3_300_3_ntrn_4996_nbat_4996_nepo_1000_PCA_ZCB_USD")
data("GMMN_dim_3_600_3_ntrn_4996_nbat_4996_nepo_1000_PCA_ZCB_USD")
data("GMMN_dim_4_100_4_ntrn_4947_nbat_4947_nepo_1000_PCA_ZCB_CAD")
data("GMMN_dim_4_300_4_ntrn_4947_nbat_4947_nepo_1000_PCA_ZCB_CAD")
data("GMMN_dim_4_600_4_ntrn_4947_nbat_4947_nepo_1000_PCA_ZCB_CAD")
data("GMMN_dim_5_100_5_ntrn_5478_nbat_5478_nepo_1000_FX_USD")
data("GMMN_dim_5_300_5_ntrn_5478_nbat_5478_nepo_1000_FX_USD")
data("GMMN_dim_5_600_5_ntrn_5478_nbat_5478_nepo_1000_FX_USD")
data("GMMN_dim_6_100_6_ntrn_5478_nbat_5478_nepo_1000_FX_GBP")
data("GMMN_dim_6_300_6_ntrn_5478_nbat_5478_nepo_1000_FX_GBP")
data("GMMN_dim_6_600_6_ntrn_5478_nbat_5478_nepo_1000_FX_GBP")

Format

GMMN_dim_2_300_2_ntrn_60000_nbat_5000_nepo_300_C_tau_0.25

raw R object representing a GMMN (input and output layer are two-dimensional, the single hidden layer is 300-dimensional) trained on 60000 pseudo-samples (with batch size 5000 and 300 epochs) from a bivariate Clayton copula (with parameter chosen such that Kendall's tau equals 0.25).

GMMN_dim_2_300_2_ntrn_60000_nbat_5000_nepo_300_C_tau_0.5

raw R object representing a GMMN (input and output layer are two-dimensional, the single hidden layer is 300-dimensional) trained on 60000 pseudo-samples (with batch size 5000 and 300 epochs) from a bivariate Clayton copula (with parameter chosen such that Kendall's tau equals 0.5).

GMMN_dim_2_300_2_ntrn_60000_nbat_5000_nepo_300_C_tau_0.75

raw R object representing a GMMN (input and output layer are two-dimensional, the single hidden layer is 300-dimensional) trained on 60000 pseudo-samples (with batch size 5000 and 300 epochs) from a bivariate Clayton copula (with parameter chosen such that Kendall's tau equals 0.75).

GMMN_dim_2_300_2_ntrn_60000_nbat_5000_nepo_300_G_tau_0.25

raw R object representing a GMMN (input and output layer are two-dimensional, the single hidden layer is 300-dimensional) trained on 60000 pseudo-samples (with batch size 5000 and 300 epochs) from a bivariate Gumbel copula (with parameter chosen such that Kendall's tau equals 0.25).

GMMN_dim_2_300_2_ntrn_60000_nbat_5000_nepo_300_G_tau_0.5

raw R object representing a GMMN (input and output layer are two-dimensional, the single hidden layer is 300-dimensional) trained on 60000 pseudo-samples (with batch size 5000 and 300 epochs) from a bivariate Gumbel copula (with parameter chosen such that Kendall's tau equals 0.5).

GMMN_dim_2_300_2_ntrn_60000_nbat_5000_nepo_300_G_tau_0.75

raw R object representing a GMMN (input and output layer are two-dimensional, the single hidden layer is 300-dimensional) trained on 60000 pseudo-samples (with batch size 5000 and 300 epochs) from a bivariate Gumbel copula (with parameter chosen such that Kendall's tau equals 0.75).

GMMN_dim_2_300_2_ntrn_60000_nbat_5000_nepo_300_eqmix_C_tau_0.5_rot90_t4_tau_0.5

raw R object representing a GMMN (input and output layer are two-dimensional, the single hidden layer is 300-dimensional) trained on 60000 pseudo-samples (with batch size 5000 and 300 epochs) from a bivariate half-half mixture of a Clayton copula (with parameter chosen such that Kendall's tau equals 0.5) and a rotated (by 90 degree) $t$ copula (with 4 degrees of freedom and correlation parameter chosen such that Kendall's tau equals 0.5).

GMMN_dim_2_300_2_ntrn_60000_nbat_5000_nepo_300_eqmix_G_tau_0.5_rot90_t4_tau_0.5

raw R object representing a GMMN (input and output layer are two-dimensional, the single hidden layer is 300-dimensional) trained on 60000 pseudo-samples (with batch size 5000 and 300 epochs) from a bivariate half-half mixture of a Gumbel copula (with parameter chosen such that Kendall's tau equals 0.5) and a rotated (by 90 degree) $t$ copula (with 4 degrees of freedom and correlation parameter chosen such that Kendall's tau equals 0.5).

GMMN_dim_2_300_2_ntrn_60000_nbat_5000_nepo_300_eqmix_MO_0.75_0.6_rot90_t4_tau_0.5

raw R object representing a GMMN (input and output layer are two-dimensional, the single hidden layer is 300-dimensional) trained on 60000 pseudo-samples (with batch size 5000 and 300 epochs) from a bivariate half-half mixture of a Marshall–Olkin copula (with alpha_1 = 0.75 and alpha_2 = 0.60) and a rotated (by 90 degree) $t$ copula (with 4 degrees of freedom and correlation parameter chosen such that Kendall's tau equals 0.5).

GMMN_dim_2_300_2_ntrn_60000_nbat_5000_nepo_300_MO_0.75_0.6

raw R object representing a GMMN (input and output layer are two-dimensional, the single hidden layer is 300-dimensional) trained on 60000 pseudo-samples (with batch size 5000 and 300 epochs) from a Marshall–Olkin copula (with alpha_1=0.75 and alpha_2=0.60).

GMMN_dim_2_300_2_ntrn_60000_nbat_5000_nepo_300_t4_tau_0.25

raw R object representing a GMMN (input and output layer are two-dimensional, the single hidden layer is 300-dimensional) trained on 60000 pseudo-samples (with batch size 5000 and 300 epochs) from a two-dimensional $t$ copula (with 4 degrees of freedom and equi-correlation parameter chosen such that Kendall's tau equals 0.25).

GMMN_dim_2_300_2_ntrn_60000_nbat_5000_nepo_300_t4_tau_0.5

raw R object representing a GMMN (input and output layer are two-dimensional, the single hidden layer is 300-dimensional) trained on 60000 pseudo-samples (with batch size 5000 and 300 epochs) from a two-dimensional $t$ copula (with 4 degrees of freedom and equi-correlation parameter chosen such that Kendall's tau equals 0.5).

GMMN_dim_2_300_2_ntrn_60000_nbat_5000_nepo_300_t4_tau_0.75

raw R object representing a GMMN (input and output layer are two-dimensional, the single hidden layer is 300-dimensional) trained on 60000 pseudo-samples (with batch size 5000 and 300 epochs) from a two-dimensional $t$ copula (with 4 degrees of freedom and equi-correlation parameter chosen such that Kendall's tau equals 0.75).

GMMN_dim_3_300_3_ntrn_60000_nbat_5000_nepo_300_NC21_tau_0.25_0.5

raw R object representing a GMMN (input and output layer are three-dimensional, the single hidden layer is 300-dimensional) trained on 60000 pseudo-samples (with batch size 5000 and 300 epochs) from a three-dimensional nested Clayton copula (with sector dimensions 2 and 1, corresponding Kendall's tau 0.5 within the first sector and Kendall's tau 0.25 between the two sectors).

GMMN_dim_3_300_3_ntrn_60000_nbat_5000_nepo_300_NG21_tau_0.25_0.5

raw R object representing a GMMN (input and output layer are three-dimensional, the single hidden layer is 300-dimensional) trained on 60000 pseudo-samples (with batch size 5000 and 300 epochs) from a three-dimensional nested Gumbel copula (with sector dimensions 2 and 1, corresponding Kendall's tau 0.5 within the first sector and Kendall's tau 0.25 between the two sectors).

GMMN_dim_5_300_5_ntrn_60000_nbat_5000_nepo_300_C_tau_0.5

raw R object representing a GMMN (input and output layer are five-dimensional, the single hidden layer is 300-dimensional) trained on 60000 pseudo-samples (with batch size 5000 and 300 epochs) from a five-dimensional Clayton copula (with parameter chosen such that Kendall's tau equals 0.5).

GMMN_dim_5_300_5_ntrn_60000_nbat_5000_nepo_300_G_tau_0.5

raw R object representing a GMMN (input and output layer are five-dimensional, the single hidden layer is 300-dimensional) trained on 60000 pseudo-samples (with batch size 5000 and 300 epochs) from a five-dimensional Gumbel copula (with parameter chosen such that Kendall's tau equals 0.5).

GMMN_dim_5_300_5_ntrn_60000_nbat_5000_nepo_300_NC23_tau_0.25_0.5_0.75

raw R object representing a GMMN (input and output layer are five-dimensional, the single hidden layer is 300-dimensional) trained on 60000 pseudo-samples (with batch size 5000 and 300 epochs) from a five-dimensional nested Clayton copula (with sector dimensions 2 and 3, corresponding Kendall's tau 0.5 and 0.75, and Kendall's tau 0.25 between the two sectors).

GMMN_dim_5_300_5_ntrn_60000_nbat_5000_nepo_300_NG23_tau_0.25_0.5_0.75

raw R object representing a GMMN (input and output layer are five-dimensional, the single hidden layer is 300-dimensional) trained on 60000 pseudo-samples (with batch size 5000 and 300 epochs) from a five-dimensional nested Gumbel copula (with sector dimensions 2 and 3, corresponding Kendall's tau 0.5 and 0.75, and Kendall's tau 0.25 between the two sectors).

GMMN_dim_5_300_5_ntrn_60000_nbat_5000_nepo_300_t4_tau_0.5

raw R object representing a GMMN (input and output layer are five-dimensional, the single hidden layer is 300-dimensional) trained on 60000 pseudo-samples (with batch size 5000 and 300 epochs) from a five-dimensional $t$ copula (with 4 degrees of freedom and equi-correlation parameter chosen such that Kendall's tau equals 0.5).

GMMN_dim_10_300_10_ntrn_60000_nbat_5000_nepo_300_C_tau_0.5

raw R object representing a GMMN (input and output layer are 10-dimensional, the single hiddenlayer is 300-dimensional) trained on 60000 pseudo-samples (with batch size 5000 and 300 epochs) from a 10-dimensional Clayton copula (with parameter chosen such that Kendall's tau equals 0.5).

GMMN_dim_10_300_10_ntrn_60000_nbat_5000_nepo_300_G_tau_0.5

raw R object representing a GMMN (input and output layer are 10-dimensional, the single hiddenlayer is 300-dimensional) trained on 60000 pseudo-samples (with batch size 5000 and 300 epochs) from a 10-dimensional Gumbel copula (with parameter chosen such that Kendall's tau equals 0.5).

GMMN_dim_10_300_10_ntrn_60000_nbat_5000_nepo_300_NC55_tau_0.25_0.5_0.75

raw R object representing a GMMN (input and output layer are 10-dimensional, the single hidden layer is 300-dimensional) trained on 60000 pseudo-samples (with batch size 5000 and 300 epochs) from a 10-dimensional nested Clayton copula (with sector dimensions 5 and 5, corresponding Kendall's tau 0.5 and 0.75, and Kendall's tau 0.25 between the two sectors).

GMMN_dim_10_300_10_ntrn_60000_nbat_5000_nepo_300_NG55_tau_0.25_0.5_0.75

raw R object representing a GMMN (input and output layer are 10-dimensional, the single hidden layer is 300-dimensional) trained on 60000 pseudo-samples (with batch size 5000 and 300 epochs) from a 10-dimensional nested Gumbel copula (with sector dimensions 5 and 5, corresponding Kendall's tau 0.5 and 0.75, and Kendall's tau 0.25 between the two sectors).

GMMN_dim_10_300_10_ntrn_60000_nbat_5000_nepo_300_t4_tau_0.5

raw R object representing a GMMN (input and output layer are 10-dimensional, the single hidden layer is 300-dimensional) trained on 60000 pseudo-samples (with batch size 5000 and 300 epochs) from a 10-dimensional $t$ copula (with 4 degrees of freedom and equi-correlation parameter chosen such that Kendall's tau equals 0.5).

GMMN_dim_3_100_3_ntrn_4996_nbat_4996_nepo_1000_PCA_ZCB_USD

raw R object representing a GMMN (input and output layer are 3-dimensional, the single hidden layer is 100-dimensional) trained on 4996 pseudo-observations (with batch size 4996 and 1000 epochs) constructed using the first 3 principal components of the standardized residuals after de-ARMA(1,1)-GARCH(1,1)-ing a 30-dimensional US ZCB yield curve series from 1995-01-01 to 2014-12-31.

GMMN_dim_3_300_3_ntrn_4996_nbat_4996_nepo_1000_PCA_ZCB_USD

raw R object representing a GMMN (input and output layer are 3-dimensional, the single hidden layer is 300-dimensional) trained on 4996 pseudo-observations (with batch size 4996 and 1000 epochs) constructed using the first 3 principal components of the standardized residuals after de-ARMA(1,1)-GARCH(1,1)-ing a 30-dimensional US ZCB yield curve series from 1995-01-01 to 2014-12-31.

GMMN_dim_3_600_3_ntrn_4996_nbat_4996_nepo_1000_PCA_ZCB_USD

raw R object representing a GMMN (input and output layer are 3-dimensional, the single hidden layer is 600-dimensional) trained on 4996 pseudo-observations (with batch size 4996 and 1000 epochs) constructed using the first 3 principal components of the standardized residuals after de-ARMA(1,1)-GARCH(1,1)-ing a 30-dimensional US ZCB yield curve series from 1995-01-01 to 2014-12-31.

GMMN_dim_4_100_4_ntrn_4947_nbat_4947_nepo_1000_PCA_ZCB_CAD

raw R object representing a GMMN (input and output layer are 4-dimensional, the single hidden layer is 100-dimensional) trained on 4947 pseudo-observations (with batch size 4947 and 1000 epochs) constructed using the three 4 principal components of the standardized residuals after de-ARMA(1,1)-GARCH(1,1)-ing a 120-dimensional CAD ZCB yield curve series from 1995-01-01 to 2014-12-31.

GMMN_dim_4_300_4_ntrn_4947_nbat_4947_nepo_1000_PCA_ZCB_CAD

raw R object representing a GMMN (input and output layer are 4-dimensional, the single hidden layer is 300-dimensional) trained on 4947 pseudo-observations (with batch size 4947 and 1000 epochs) constructed using the first 4 principal components of the standardized residuals after de-ARMA(1,1)-GARCH(1,1)-ing a 120-dimensional CAD ZCB yield curve series from 1995-01-01 to 2014-12-31.

GMMN_dim_4_600_4_ntrn_4947_nbat_4947_nepo_1000_PCA_ZCB_CAD

raw R object representing a GMMN (input and output layer are 4-dimensional, the single hidden layer is 600-dimensional) trained on 4947 pseudo-observations (with batch size 4947 and 1000 epochs) constructed using the first 4 principal components of the standardized residuals after de-ARMA(1,1)-GARCH(1,1)-ing a 120-dimensional CAD ZCB yield curve series from 1995-01-01 to 2014-12-31.

GMMN_dim_5_100_5_ntrn_5478_nbat_5478_nepo_1000_FX_USD

raw R object representing a GMMN (input and output layer are 5-dimensional, the single hidden layer is 100-dimensional) trained on 5478 pseudo-observations (with batch size 5478 and 1000 epochs) constructed using the standardized residuals after de-ARMA(1,1)-GARCH(1,1)-ing a 5-dimensional USD FX return series from 2000-01-01 to 2014-12-31.

GMMN_dim_5_300_5_ntrn_5478_nbat_5478_nepo_1000_FX_USD

raw R object representing a GMMN (input and output layer are 5-dimensional, the single hidden layer is 300-dimensional) trained on 5478 pseudo-observations (with batch size 5478 and 1000 epochs) constructed using the standardized residuals after de-ARMA(1,1)-GARCH(1,1)-ing a 5-dimensional USD FX return series from 2000-01-01 to 2014-12-31.

GMMN_dim_5_600_5_ntrn_5478_nbat_5478_nepo_1000_FX_USD

raw R object representing a GMMN (input and output layer are 5-dimensional, the single hidden layer is 600-dimensional) trained on 5478 pseudo-observations (with batch size 5478 and 1000 epochs) constructed using the standardized residuals after de-ARMA(1,1)-GARCH(1,1)-ing a 5-dimensional USD FX return series from 2000-01-01 to 2014-12-31.

GMMN_dim_6_100_6_ntrn_5478_nbat_5478_nepo_1000_FX_GBP

raw R object representing a GMMN (input and output layer are 6-dimensional, the single hidden layer is 100-dimensional) trained on 5478 pseudo-observations (with batch size 5478 and 1000 epochs) constructed using the standardized residuals after de-ARMA(1,1)-GARCH(1,1)-ing a 6-dimensional GBP FX return series from 2000-01-01 to 2014-12-31.

GMMN_dim_6_300_6_ntrn_5478_nbat_5478_nepo_1000_FX_GBP

raw R object representing a GMMN (input and output layer are 6-dimensional, the single hidden layer is 300-dimensional) trained on 5478 pseudo-observations (with batch size 5478 and 1000 epochs) constructed using the standardized residuals after de-ARMA(1,1)-GARCH(1,1)-ing a 6-dimensional GBP FX return series from 2000-01-01 to 2014-12-31.

GMMN_dim_6_600_6_ntrn_5478_nbat_5478_nepo_1000_FX_GBP

raw R object representing a GMMN (input and output layer are 6-dimensional, the single hidden layer is 600-dimensional) trained on 5478 pseudo-observations (with batch size 5478 and 1000 epochs) constructed using the standardized residuals after de-ARMA(1,1)-GARCH(1,1)-ing a 6-dimensional GBP FX return series from 2000-01-01 to 2014-12-31.

Author(s)

Marius Hofert and Avinash Prasad

Source

GPU server with NVIDIA Tesla P100 GPUs.

References

Hofert, M., Prasad, A. and Zhu, M. (2018). Quasi-Monte Carlo for multivariate distributions via generative neural networks. (See https://arxiv.org/abs/1811.00683 for an early version) Hofert, M. Prasad, A. and Zhu, M. (2019). Multivariate time-series modeling with generative neural networks (See arXiv preprint arXiv:2002.10645 for an early version)

See Also

GMMN_model(), to_callable()

Examples

 # to avoid win-builder error "Error: Installation of TensorFlow not found"
## Load a trained GMMN (see train_once())
NNname <- "GMMN_dim_2_300_2_ntrn_60000_nbat_5000_nepo_300_eqmix_C_tau_0.5_rot90_t4_tau_0.5"
NN <- read_rda(NNname, package = "gnn")
GMMN1 <- to_callable(NN)
str(GMMN1)

## Alternative
NNnm <- data(list = NNname)
GMMN2 <- to_callable(get(NNnm))
str(GMMN2)

## Check (the check-able components)
stopifnot(identical(GMMN1[names(GMMN1) != "model"],
                    GMMN2[names(GMMN2) != "model"]))

## Evaluate
set.seed(271)
N.prior <- matrix(rnorm(2000 * 2), ncol = 2)
X <- predict(GMMN1[["model"]], x = N.prior)
plot(X, xlab = expression(X[1]), ylab = expression(X[2]))


[Package gnn version 0.0-3 Index]