What is Time Series Analysis?
Table of contents
Time Series Analysis
Time series analysis can be defined as the following:
Extracting meaningful summary and statistic from points arranged in chronological order
Analysis goes both ways
- Diagnose past events
- Predict future events
Analysis of time series rose from various disciplines and applications:
- Medicine
- Weather
- Economics
- Astronomy
ECG and EEG are most common time series data in medicine. In the past, these fields faced challenges due to the scarcity of recordings and bias of data towards those with symptoms. However, with the advancement of wearable technology, it is now possible to collect routine measurements.
Time series analysis have developed tremendously with the advancement of computing power.
Early Time Series Analysis
Traditional analysis models often presupposed its own outcome. For example, a cyclical model presupposes cyclical data. Which results in a lot of shortcomings.
With the emergence of autoregressive models, time series analysis became more flexible.
Unlike some other fields that are driven by theory, time series analysis was driven by practical applications.
- Businesses led the field because they had more data than academics
- Practical results were more important than underlying theory