Isolationforestprocessor

Promoting isolationforestprocessor near realtime anomaly detection processor to alpha

Description

Adding the initial implementation of the isolationforestprocessor for OpenTelemetry Collector.
This feature introduces anomaly detection using the Online Isolation Forest algorithm applied to streaming - traces, metrics, and logs. The processor evaluates each incoming signal and tags it with an anomaly_score and anomaly_is_anomaly attribute based on a configurable threshold and contamination rate using a moving time window enabling us to perform anomaly detection in near-realtime instead of performing the same as an async analysis increasing the MTTD.

This lays the foundation for unsupervised anomaly detection in observability pipelines.

Link to tracking issue

New component: Isolation-Forest anomaly detection Processor #41707