Learn R Programming

RegimeChange (version 0.1.1)

transformer_detect: Transformer-based Changepoint Detection

Description

Implements a transformer architecture inspired by TCDformer for time series changepoint detection using self-attention mechanisms.

Usage

transformer_detect(
  data,
  true_changepoints = NULL,
  window_size = 128,
  d_model = 64,
  n_heads = 4,
  n_layers = 2,
  d_ff = 256,
  dropout = 0.1,
  epochs = 50,
  threshold = 0.5,
  verbose = FALSE
)

Value

List with changepoints, attention weights, and model

Arguments

data

Numeric vector of time series data

true_changepoints

Optional vector of known changepoints

window_size

Input window size (default: 128)

d_model

Model dimension (default: 64)

n_heads

Number of attention heads (default: 4)

n_layers

Number of transformer layers (default: 2)

d_ff

Feed-forward dimension (default: 256)

dropout

Dropout rate (default: 0.1)

epochs

Training epochs (default: 50)

threshold

Detection threshold (default: 0.5)

verbose

Show progress (default: FALSE)

References

Wu, H., et al. (2023). TimesNet: Temporal 2D-Variation Modeling

Zhou, H., et al. (2021). Informer: Efficient Transformer for Long Sequence Time-Series Forecasting