In the era of big data and advanced astronomical surveys, time domain astronomy has become a cornerstone for understanding dynamic processes in the universe. We describe varieties of time series in sub-fields of astronomy and other fields to bring out their different needs.
We then go into traditional and innovative machine learning methodologies -- including Foundation models from industry -- and tools employed to analyze time series data.
On the data side we will begin with an overview of the Zwicky Transient Facility (ZTF) and its contributions to capturing irregular time series data of transient events such as supernovae. We will highlight the unique operational modes of ZTF, including "Read While Expose" and deep drilling fields, which enhance its observational capabilities. We will describe SCoPe — an advanced feature extraction method for irregular time series — and its application to ZTF.
Finally we will describe some other datasets including future ones and show the connectivity between irregular and regular light curves, and traditional and new methods including gap analysis and connect with population studies and all important interpretibility and explainability.