Time Series

A time series is a sequence of measurements over time, at equal time intervals:

• Stock market
• Meteorology
• Sales
• ECG data
• Patient vitals
• Lab values
• Epidemics
• Engineering (digital signal processing)

What can you do with time series?

• Similarity search: based on a time series find the most similar time series in a dataset
• Anomaly detection: find the subsequences of a time series that contain anomalies / unexpected behavior based on a model of “normal” behavior
• Frequent pattern mining: extracting frequent itemsets among all the time series in a dataset
• Clustering: find groupings of all the time series in a dataset
• Classification: identify the class that a time series belongs to based on multiple labeled time series in a dataset
• Regression: determine the relationaship between two time series
• Prediction: predict the next value of a time series from the values of one or multiple time series in a dataset.