An R package implementing the GARCH-Informed Neural Network (GINN) hybrid framework for volatility (variance) forecasting.
Step 1: y_t → r_t = (y_t - y_{t-1}) / y_{t-1} [returns]
Step 2: AR(p) on r_t → μ̂_t [mean return]
Step 3: GARCH(p,q) on r_t → σ²̂_GARCH [conditional variance]
Step 4: σ²_t = (r_t - μ̂_t)² [ground-truth variance]
Step 5: LSTM trained on σ²_t sequences with GINN loss:
Loss = λ × MSE(σ²_t, σ²̂_LSTM) [vs ground truth]
+ (1-λ) × MSE(σ²̂_GARCH, σ²̂_LSTM) [vs GARCH — Eq. 17]Lambda interpretation: - λ = 1.0 →
Standard LSTM (ground-truth variance only, no GARCH guidance) -
λ = 0.5 → Balanced GINN (equal weight) -
λ = 0.0 → GINN-0 (LSTM only learns from GARCH)
install.packages(c("torch", "rugarch", "ggplot2", "cli", "coro", "devtools"))
torch::install_torch() # one-time ~500 MB download
devtools::install("path/to/GINN")library(GINN)
# Simulated prices
prices <- cumprod(1 + rnorm(150, 0.001, 0.02)) * 100
# ── AUTO mode ─────────────────────────────────────────────────
# AR lag, GARCH config, LSTM hyperparameters all auto-selected
result <- GINN(
data = prices,
mode = "auto",
seq_len = 5,
lambda_list = seq(0.1, 0.9, 0.1)
)
print(result)
plotGINN(result, h = 12)
predictGINN(result, h = 12)
plotlambda(result)
accuracy.GINN(result) # this one stays # full metrics table
# ── MANUAL mode ───────────────────────────────────────────────
result_m <- GINN(
data = prices,
mode = "manual",
ar_lag = 1,
garch_p = 1, garch_q = 1,
garch_mean = "zero", garch_dist = "norm",
seq_len = 5, hidden_size = 32,
num_layers = 1, lr = 0.001, dropout = 0.0,
lambda_list = c(0.1, 0.5, 0.9)
)
print(result_m)| Parameter | Default | Description |
|---|---|---|
data |
required | Numeric price/yield vector |
mode |
"auto" |
"auto" or "manual" |
ar_max_lag |
5 |
Max AR lag to search (auto) |
ar_lag |
NULL |
Fixed AR lag (manual) |
garch_p_grid |
1:3 |
ARCH order grid (auto) |
garch_q_grid |
1:3 |
GARCH order grid (auto) |
garch_mean_grid |
c("zero","arma") |
Mean model grid (auto) |
garch_dist_grid |
c("norm","std") |
Distribution grid (auto) |
garch_p/q/mean/dist |
NULL |
Fixed GARCH config (manual) |
seq_len |
5 |
Variance sequence length |
hidden_size |
32 |
LSTM hidden units (manual) |
num_layers |
1 |
Stacked layers (manual) |
lr |
0.001 |
Learning rate (manual) |
dropout |
0.0 |
Dropout rate (manual) |
epochs |
500 |
Max training epochs |
patience |
30 |
Early-stopping patience |
tune_epochs |
50 |
Epochs per grid combo (auto) |
batch_size |
16 |
Mini-batch size |
lambda_list |
0.1…0.9 |
Lambda values to test |
hidden_grid |
c(8,16,24,32,40) |
Hidden grid (auto) |
layers_grid |
c(1,2) |
Layers grid (auto) |
lr_grid |
c(0.001,0.003,0.005) |
LR grid (auto) |
dropout_grid |
c(0.0,0.1,0.2,0.3) |
Dropout grid (auto) |
GINN
class)result$results # list — one per lambda (RMSE, MAE, R², preds, loss_hist)
result$garch # GARCH config, coefs, sigma²_train/test, metrics
result$ar # AR lag, fitted values, forecast
result$returns # computed return series r_t
result$variance # ground-truth variance σ²_t (train/test/all)
result$best_model # complete best model summary
result$best_hp # best LSTM hyperparameters
result$tuning_log # grid-search log (auto mode)
result$meta # settings
result$data_info # original data, split, Nprint(result) # compact table (all lambdas)
summary(result) # full data frame with all metrics
plotGINN(result, h = 12) # actual + predicted + forecast
plotGINN(result, h = 12, actual = new_prices) # with actual future data
plotlambda(result) # Training RMSE bar chart
predictGINN(result, h = 12) # variance forecast
AccuracyGINN(result, actual = new_prices) # accuracy vs actual
plot(result, "loss") # training loss (unchanged)
accuracy.GINN(result) # accuracy data frametorch — LSTM (native R, no Python required)rugarch — GARCH fittingggplot2 — plotscli — progress messagescoro — dataloader iteration