A research team has collected a large dataset of sensor readings from various industrial machines.

A research team has collected a large dataset of sensor readings from various industrial machines. This dataset includes measurements like temperature, pressure, vibration levels, and electrical current, recorded at regular intervals. The team has not yet assigned any labels or categories to these readings and wants to identify potential anomalies, malfunctions, or natural groupings of machine behavior based on the sensor data alone. What type of machine learning should they use?