Time Series Analysis and Mining Techniques

Abstract

A time series is a sequence of data values that are recorded with equal or varying time intervals. Time series data usually includes timestamps that indicate the time at which each individual value in the times series is recorded. Time series data is usually transmitted in the form of a data stream, i.e., continuous flow of data values. Source of time series data can be any system that measures and records data values over the course of time. Some examples of time series data may be recorded from stock values, blood pressure of a patient, temperature of a room, amount of a product in the inventory, and amount of precipitation in a region. Proper analysis and mining of time series data may yield valuable knowledge about the underlying characteristics of the data source. Time series analysis and mining has applications in many domains, such as financial, biomedical, and meteorological applications, because time series data may be generated by various sources in different domains.