Docs

Torch Log Service

Machine learning function

Overview of machine learning functions

Overview of machine learning functions

This topic lists the machine learning functions supported by TLS.

Notice

Currently, machine learning functions are in public beta.

Prediction and anomaly detection functions

Function	Description
ts_anomaly_filter	Quickly identifies anomaly curves by filtering anomaly detection results with custom exception patterns.
ts_predicate_simple	Completes model data and performs forecasting and anomaly detection using default parameters.
ts_predicate_ar	Completes model data and conducts predictions and anomaly detection with autoregressive models.
ts_predicate_arima	Conducts accurate prediction on time series data with periodic characteristics.
ts_regression_predict	Filter anomaly detection results based on defined anomaly patterns to quickly identify exception curves.

Time series clustering functions

Function	Description
ts_density_cluster	Use the density clustering method to cluster multiple time series curves.
ts_hierarchical_cluster	Cluster multiple time series curves using hierarchical clustering methods.
ts_similar_instance	Search time series curve that is similar to the specified time series curve.

Anomaly comparison function

Function	Description
anomaly_compare	Compare the differences of the target observation object between two time periods.

Threshold recommendation function

Function	Description
ts_threshold_recommendation	Cluster historical time series data to quickly identify the distribution of normal data.

Difference pattern statistical function

Function	Description
pattern_diff	Based on given field samples and judging criteria, analyze the set of differential patterns that affect the specified criteria.

Last updated: 2026.05.07 15:33:48