Summarizing Time Series: Learning Patterns in 'Volatile' Series

Most financial time series processes are nonstationary and their frequency characteristics are time-dependant. In this paper we present a time series summarization and prediction framework to analyse volatile and high-frequency time series data. Multiscale wavelet analysis is used to separate out the trend, cyclical fluctuations and autocorrelational effects. The framework can generate verbal signals to describe each effect. The summary output is used to reason about the future behaviour of the time series and to give a prediction. Experiments on the intra-day European currency spot exchange rates (over 10 data points per second, 24 hours per day) are described. The results are compared with a neural network prediction framework.

• Type: Conference
• Year: 2004
• Where: IDEAL
• Authors: Saif Ahmad, Tugba Taskaya-Temizel, Khurshid Ahmad
• Keywords: IDEAL 2004, Prediction Framework, Time Series, Time Series Summarization