Methods of Nonlinear Time Series Cycle Analysis in Big Data Environment and IoT Application

Abstract

There have been a lot of changes in nonlinear time series analysis techniques over the last few decades. They came from their time series background and have grown some of these techniques to try to fill in a gap in the ability to model and predict certain types of data, like chaotic and fixed systems. It is possible to find a lot of these systems all over different places in both natural and human worlds. This study explains how these techniques came to be, what they are based on, and how some of them can predict what will happen in the future. This study is trying to figure out how they work and what they could be used for. The contribution of this study is a different way to look at time series data. It looks at how the data is made and how big the time series is. Here, we talk about how the model comes together in terms of order. In this example, we look at the time series of the sectorial indices on the National Stock Exchange (India) to show that the model is getting close. Convergence also points out that, when we fit the same number of data points to the same model, we get the same thing every time. This shows that a true model correctly predicted the value of the future and that data always comes together in nature. A study found that the size of a big data time series should be taken into account when looking at the time series’ realization. This means that the DGP intervals for each study period can be the best representative intervals for that time period.