Interpolation of financial time series data in a three-dimensional spatialisation

Loading...
Thumbnail Image
Other Title
Authors
Borna, Kambiz.
Moore, A.B.
Author ORCID Profiles (clickable)
Degree
Grantor
Date
2022-08-28
Supervisors
Type
Conference Contribution - Oral Presentation
Ngā Upoko Tukutuku (Māori subject headings)
Keyword
Bitcoin, financial time series data, time series data, financial markets, spatialisation, algorithms
Bitcoin
financial time series data
time series data
financial markets
spacialisation
algorithms
ANZSRC Field of Research Code (2020)
Citation
Borna, K. & Moore, A.B (2022, August, 29-30). interpolation of financial time series data in a three-dimensional spatialisation [Paper presentation]. New Zealand Geospatial Research Conference (NZGRC) 2022, Wellington. https://hdl.handle.net/10652/6150
Abstract
This paper introduces a new approach to interpolating and visualising financial time series data, e.g., Bitcoin prices, in a spatial domain using the notion of spatialisation: forming a spatial representation of non-spatial phenomena. The proposed algorithm first utilises the temporal components of the Bitcoin prices, i.e., date and time of day, to build a 2D vector map based on four observations per day, i.e., opening, high, low and closing prices. It then uses the coordinates assigned to the prices and their values to convert the 2D vector map into a 3D raster map. This transformation is performed using the Natural Neighbour Interpolation (NNI) algorithm. We then apply the 3D map to interpolate time series data with a 30-minute frequency and compare it with the actual observed data to assess the quality of the map. The RMSE results show an improvement of 12% using the proposed method compared with conventional interpolation methods
Publisher
Link to ePress publication
DOI
Copyright holder
Authors
Copyright notice
Attribution 4.0 International
Available online at