Using Ordinal Partition Transition Networks to Analyze the Consumer Price Index
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
The Consumer Price Index (CPI) is a key economic indicator, yet traditional forecasting methods often struggle to capture its nonlinear, complex behavior. This paper applies Ordinal Partition Transition Networks (OPTNs) to Ecuador’s CPI from January 2001 to December 2024. By converting the CPI data into Markov chains through ordinal partitions, we construct transition matrices that reveal underlying dynamics and use Ordinal Permutation Entropy, Ordinal Transition Entropy, and Kullback–Leibler Divergence to measure complexity and detect outliers. We bridge a gap in economic time series analysis by integrating high-level complexity metrics and sequence-specific alignment within a single network-based framework. In systematically examining 480 model configurations, we find that smaller embedding dimensions yield more accurate CPI forecasts with high entropy values and moderate divergence, whereas higher dimensions exhibit stronger sensitivity to out-of-distribution events. Thus, OPTNs capture predictive structure and highlight rare behaviors, providing policymakers and analysts with a robust tool for uncovering intricate economic patterns and guiding data-driven decisions.
Fields of Study & Subjects
Complex Systems and Time Series Analysis
Time Series Analysis and Forecasting
Chaos control and synchronization
Publication details
- Journal / source
- Revista Politécnica
- Volume
- 57
- Issue
- 1
- Pages
- 109–122
- Publication date
- 28 February 2026
- Language
- EN