Top Page | Upper Page | Contents | About This Site | JAPANESE

Time Series Analysis

Time Series Analysis is used to analyze changing condition. It is the analysis of dynamics.

Usually, the images of time series are movement of temperature or stocks. But, for example, analysis of log-data of machines and MES (Manufacturing Execution System) is the kinds of time series analysis.

Most data related with time. So the idea of time series analysis is important in Data Science.

Data Literacy

Data Literacy for time data is below.

• Line Graph : Basic and useful tool
• Treatment of Time : Computers deal with time data in special way.
• Velocity Data : Easy but useful
• 0-1 Data Analysis : Direct analysis is not good.

Kinds of Time Series Analysis

Sensor Data Analysis needs various aproach.

• Condision Analysis : Pre-treatment of data is not much.
• Time Analysis : Analysis of "timing" and "length of period".
• Passing Analysis : Analysis of the change from start to end.
• Self Correlation Analysis : Analysis of difference of steps.
• Moving Analysis : Simple analysis for complex relationship
• Spectrum Analysis : Analysis of periodicity and frequency.
• Additive model and multiplicative model : Expressing increased variability

Contents of "Time Series Analysis"

When you find the title, "Time Series Analysis", it may deal with Self Correlation Analysis or Spectrum Analysis .

NEXT Condition Analysis