Until recently, the use of time series data was mainly related to fields of science, such as economics, finance, physics, engineering, and astronomy. However, in recent years, as the ability to collect data improved with the use of digital devices such as computers, mobiles, sensors, or satellites, time series data is now everywhere. The enormous amount of data that's collected every day probably goes beyond our ability to observe, analyze, and understand it.
The development of time series analysis and forecasting did not start with the introduction of the stochastic process during the previous century. Ancient civilizations such as the Greeks, Romans, or Mayans researched and learned how to utilize cycled events such as weather, agriculture, and astronomy over time to forecast future events. For example, during the classic period of the Mayan civilization (between 250 AD and 900 AD), the Maya priesthood assumed that there are cycles in astronomy events and therefore they patiently observed, recorded, and learned those events. This allowed them to create a detailed time series table of past events, which eventually allowed them to forecast future events, such as the phases of the moon, eclipses of the moon and the sun, and the movement of stars such as Venus, Jupiter, Saturn, and Mars. The Mayan's priesthood used to collect data and analyze the data to identify patterns and cycles. This analysis was then utilized to predict future events. We can find a similarity between the Mayan's ancient analytical process and the time series analysis process we use now. However, the modern time series analysis process is based on statistical modeling and heavy calculations that are possible with today's computers and software, such as R.
Now that we defined the main characteristics of time series data, we can move forward and start to discuss the main characteristics of time series analysis.
