论文标题
关于时间序列的先验预测和简单性偏见的注释
A Note on A Priori Forecasting and Simplicity Bias in Time Series
论文作者
论文摘要
我们可以在多大程度上预测一个时间序列而不适合历史数据？通用模式的概率模式可以在此任务中有所帮助吗？最近使用了模式的Kolmogorov复杂性与模式概率之间的深层关系，以使物理，生物学和工程系统的各种系统中的概率预测\ Emph {emph {emph {emph {emph {emph {emph {a。在这里，我们研究\ emph {简单性偏见}（SB） - 模式概率的指数上限衰减，复杂性的增加 - 在从世界银行开放数据收集中提取的离散时间序列中。我们预测了离散串联模式的概率上的上限，而不适合数据趋势。因此，我们执行一种“没有训练数据的预测”，预测时间序列形状模式\ emph {a priviri}，而不是该系列的实际数值。此外，我们可以通过使用每个系列的复杂性，对两个离散级数中的哪个更有可能更可能具有$ \ sim $ 80 \％的准确性，远高于50 \％的基线速率。这些结果表明，实用时间序列的预测和与机器学习方法的集成有前途的观点。
To what extent can we forecast a time series without fitting to historical data? Can universal patterns of probability help in this task? Deep relations between pattern Kolmogorov complexity and pattern probability have recently been used to make \emph{a priori} probability predictions in a variety of systems in physics, biology and engineering. Here we study \emph{simplicity bias} (SB) -- an exponential upper bound decay in pattern probability with increasing complexity -- in discretised time series extracted from the World Bank Open Data collection. We predict upper bounds on the probability of discretised series patterns, without fitting to trends in the data. Thus we perform a kind of `forecasting without training data', predicting time series shape patterns \emph{a priori}, but not the actual numerical value of the series. Additionally we make predictions about which of two discretised series is more likely with accuracy of $\sim$80\%, much higher than a 50\% baseline rate, just by using the complexity of each series. These results point to a promising perspective on practical time series forecasting and integration with machine learning methods.
