Data analysis support by storing and evaluating time series-based information

OData support
Supervisor:
Dr. Gönczy László
Department of Measurement and Information Systems

The data driven programs have an ever growing importance in our lives, especially in the environment of businesses, multinational companies. One of the most important part of this is to properly secure and store these data. In my dissertation I’ll explore time series based storage tools, which were chosen as the four best time series based databases in 2019. I present the tools necessary to display the data and the qualities of time based series data.

I'll present the consept time series data, and the methods to analyze these data. To explore the dependency between data I present linear regression and correlation. For time based data I’ll demonstrate Arima, which helps to create predictions, and also able to calculate moving average.

I’ll present my own application, which saves data to a time based Influxdb. My software helps to identify the quality of the data sets. There’s an option with the saved data, that allows us to test the dependancy of the data columns of our data sets focusing on linear regression and correlation. The application is capable doing Arima calculations, which answer prediction mistake proportions, and also a prediction based forecast. It is also capable of calculating moving average based ont he Arima paramters.

At the end of my dissertation I’ll demonstrate the operation of my program with data about ten years of temperature in Szeged.

Downloads

Please sign in to download the files of this thesis.