Histogram Time Series Forecasting: An Application in Finance

Histogram time series describe situations where a distribution of values is available for each instant of time. These situations usually arise when contemporaneous or temporal aggregation is required. In these cases, histograms provide a summary of the data that is more informative than those provided by other aggregates such as the mean. The presentation will show two different approaches to forecast histogram time series. Namely, exponential smoothing methods and the k-Nearest Neighbours (k-NN) algorithm. Both approaches are based on the concept of barycentric histogram. The barycentric histogram emulates the “average” operation, which is the key to adapt the smoothing filters and the k-NN. The methods proposed will be illustrated with the help of a time series of daily histograms that aggregate the 5-min intra-daily returns.

This event is organised by the Econometric Institute.
Twitter: @MetricsSeminars