US 11,789,977 B1
Hierarchical aggregation and disaggreation of time series data forecasts
Chetan Mehta, Seattle, WA (US); Anish Borkar, Cambridge, MA (US); Parnika Singh, Seattle, WA (US); Divya Hariharan, Seattle, WA (US); Anup Bharadwaj, Seattle, WA (US); and Yasaswi Vempati, Seattle, WA (US)
Assigned to Amazon Technologies, Inc., Seattle, WA (US)
Filed by Amazon Technologies, Inc., Seattle, WA (US)
Filed on Sep. 27, 2021, as Appl. No. 17/486,599.
Int. Cl. G06F 16/28 (2019.01); G06F 16/22 (2019.01); G06F 16/2458 (2019.01)
CPC G06F 16/282 (2019.01) [G06F 16/2264 (2019.01); G06F 16/2477 (2019.01); G06F 16/287 (2019.01)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method, comprising:
obtaining, at a demand planning service, a time series dataset, wherein the time series dataset includes a set of dimensions and a plurality of hierarchy levels where dimensions of the set of dimensions are associated with a particular hierarchy level of the plurality of hierarchy levels;
generating forecasts of values for the set of dimensions based at least in part on the time series dataset to generate a forecasted time series dataset;
storing, in a data store, the forecasted time series dataset for display to a user of the demand planning service;
obtaining, at an interface of the demand planning service, an override value for a value of a first dimension of the set of dimensions of the forecasted time series dataset, the first dimension associated with a first hierarchy level being at least one level higher than a second hierarchy level;
determining a set of proportions based at least in part on values of a subset of dimensions of the set of dimensions associated with the second hierarchy level relative to a total of the values of the subset of dimensions;
disaggregating the override value to the values of the subset of dimensions based at least in part on the set of proportions to generate an override time series dataset including the plurality of hierarchy levels; and
updating the data store to include the override time series data set.