Greenhouse: A Zero-Positive Machine Learning System for Time-Series Anomaly Detection

This addresses anomaly detection in time-series data, but it appears incremental as it builds on zero-positive learning concepts without claiming major breakthroughs.

The paper tackles the problem of time-series anomaly detection by proposing Greenhouse, a zero-positive machine learning system, but no concrete results or numbers are provided as it describes ongoing research.