A1090
Title: Adjusting for seasonality using point process models
Authors: Anindya Roy - U.S. Census Bureau (United States) [presenting]
Tucker McElroy - Census Bureau (United States)
Abstract: Data publishing agencies are increasingly publishing data collected at a higher frequency than in the past. This has resulted in several published mixed-frequency time series data. Seasonal adjustment of mixed frequency time series is a challenging problem. We use a marked point process model to model the series as aggregation at different frequency levels. We use a transformation of the intensity measure to excise periodic components, thereby adjusting the data to be non-seasonal. The method is illustrated with daily, weekly, and monthly time series.
