A Review on Topological Data Analysis in Time Series

In data science, a time series is a predefined method of examining a set of data points gathered over time. A considerable amount of data is required for time series analysis to guarantee dependability and consistency. Most time series have both chaotic and non-chaotic components, making it important to isolate the crucial elements. It follows that there must be a unique method for extracting such characteristics from highly dynamic time series. Researchers have used a novel approach that was previously thought to have no bearing on time series. The coordinate-free and variance-free characteristics of topological data analysis are one reason why it is becoming more popular. To capture the structure of data, topological data analysis uses mathematical methods. Time series with a lot of noise have been managed with the use of topological data analysis. Topological data analysis uses persistent homology as its foundational methodology, which is then used in the study of time series and signal processing. This paper gives a review of topological approach towards time series processing and pipeline of it. In the application, we have reviewed one application related to room occupancy data using multivariate time series. With the Topological approach, they have obtained better results as compared to traditional methods. In future, the application of Topological approach towards time series application will give accurate predictions.