Hypergraph neural network classifier (HGNN)

Trains the two-layer hypergraph convolutional network of Feng et al. (2019) on a hypergraph: each layer propagates linearly transformed vertex features through \(G = D_v^{-1/2} H W D_e^{-1} H^T D_v^{-1/2}\) (the Zhou similarity operator), with ReLU and dropout between layers and a cross-entropy loss on the labeled vertices. Needs the suggested torch package.

Usage

hg_neural(
  hg,
  labels,
  features = "incidence",
  hidden = 128L,
  epochs = 600L,
  lr = 0.01,
  weight_decay = 5e-04,
  dropout = 0.5,
  validation = 0.1,
  edge_weights = NULL,
  seed = 1L,
  verbose = FALSE
)

Arguments

hg: A text_hypergraph() (or any hypernets net_hg), dense or sparse.
labels: The known labels: a named character vector (names are node identifiers, values class labels) or a tidy data.frame with a node column and a label, cluster or predicted column. At least two classes.
features: Vertex feature matrix, one row per node in hg node order (rownames must match the node names), e.g. sbert embeddings. The default "incidence" uses the hypergraph's own weighted incidence rows (tf-idf bag-of-words features for a nodes = "doc" text hypergraph).
hidden: Width of the hidden layer.
epochs: Training epochs (Adam).
lr, weight_decay: Adam learning rate and L2 penalty. The defaults (lr = 0.01, epochs = 600) were selected by validation accuracy on held-out seeds for high-dimensional sparse text features; the original paper's lr = 0.001, 200 epochs underfits that regime.
dropout: Dropout rate after the hidden layer.
validation: Fraction of the labeled seeds held out (stratified) to pick the best epoch; 0 trains on all seeds for epochs and keeps the final weights.
edge_weights: Optional positive hyperedge weights W (one per hyperedge); default all 1, the paper's initialization.
seed: Integer seed for weight initialization, dropout and the validation split (torch and R RNGs); results are deterministic given a seed.
verbose: If TRUE, message the loss every 20 epochs.

Value

A base data.frame, one row per node, with columns node, label (the given label or NA), predicted, score (softmax probability of the winning class) and margin (winner minus runner-up probability). The training history is attached as attribute "history" (data.frame: epoch, loss, val_accuracy).

References

Feng, Y., You, H., Zhang, Z., Ji, R., & Gao, Y. (2019). Hypergraph neural networks. AAAI 33.

Examples

# \donttest{
if (requireNamespace("torch", quietly = TRUE)) {
  hg <- text_hypergraph(c(
    cooking_1 = "simmer the soup with onions and carrots",
    cooking_2 = "this soup recipe needs salt on a cold night",
    space_1 = "the telescope revealed a distant galaxy and stars",
    space_2 = "astronomers aimed the telescope at the stars all night"
  ), stop_words = c("the", "with", "and", "a", "this", "at", "on", "all"))
  hg_neural(hg, labels = c(cooking_1 = "cooking", space_1 = "space"),
            hidden = 8, epochs = 50, validation = 0)
}
# }