Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].
The Arrow of Time in Multivariate Time Series
Authors: Stefan Bauer, Bernhard Schölkopf, Jonas Peters
ICML 2016 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We present a slightly modified practical algorithm that estimates the time direction for a given sample and prove its consistency. We further investigate how the performance of the algorithm depends on sample size, number of dimensions of the time series and the order of the process. An application to real world data from economics shows that considering multivariate processes instead of univariate processes can be beneficial for estimating the time direction. Our result extends earlier work on univariate time series. |
| Researcher Affiliation | Academia | Stefan Bauer EMAIL Department of Computer Science, ETH Zurich 8092 Zurich, Switzerland Bernhard Sch olkopf EMAIL Jonas Peters EMAIL Max Planck Institute for Intelligent Systems, 72076 T ubingen, Germany |
| Pseudocode | Yes | Algorithm 1 Detecting the direction of multivariate time series |
| Open Source Code | Yes | The code is available as supplementary material. |
| Open Datasets | Yes | In a dataset containing the quarterly growth rates of real gross domestic product (GDP) of UK, Canada and USA from 1980 to 2011 (Tsay, 2014), we tested our approach for different time lags, see Figure 1. |
| Dataset Splits | No | The paper mentions 'significance levels of sig1 = 0.1 and sig2 = 0.05' for the algorithm's decision process but does not provide details on training, validation, or test dataset splits. |
| Hardware Specification | No | The paper mentions using 'the Econometrics Toolbox within Matlab' but does not specify any hardware details like CPU, GPU models, or memory used for the experiments. |
| Software Dependencies | No | For simulating (function vgxproc) and fitting (function vgxvarx) VAR(p) processes we used the Econometrics Toolbox within Matlab, which in turn uses maximum likelihood to estimate the parameters. |
| Experiment Setup | Yes | For all experiments, we use significance levels of sig1 = 0.1 and sig2 = 0.05. This is a conservative but interpretable choice that could be changed in order to increase the performance in the simulations. |