Deep Learning–Based Changepoint Detection with Local Neural Models

Documentation for package ‘scanCP’ version 0.1.0

best_split_free	Best Split for Binary Labels (Free Assignment)
calc_detector	Compute the MLP-Based Changepoint Detector Statistic
combine_detectors	Combine Multiple Detector Statistics into a Joint Detector
decompose_signal_core	Core Signal Decomposition (Changepoints + Piecewise Correction)
detect_cp_ecdf	ECDF-Based Changepoint Detection for a Detector Statistic
fit_global_mlp	Fit Global MLP and Compute Residual Diagnostics
fit_mlp	Fit Rolling MLP Models for Change Point Detection
ma	Two-Sided Moving Average Smoothing
plot_mlp_fits_interactive	Interactive Plot of Rolling MLP Fits
scan_cp	Full Changepoint Detection Pipeline Using Rolling MLPs
scan_cp_multi_async	Asynchronous Multivariate Changepoint Detection
scan_cp_multi_sync	Multivariate Synchronized Changepoint Detection
select_best_spike	Select the Most Significant Spike in an ECDF Spacing Curve
simulate_piecewise_signal_idx	Simulate a Piecewise-Constant Signal with Smooth Trend and Noise