N-BEATS is a neural-network based model for univariate timeseries forecasting. N-BEATS is a ServiceNow Research project that was started at Element AI.
In the MASE, I find that the shapes of in_sample, out_sample, and forecasting should be time_o or time_in, instead of batch, time_i/o.
Moreover, does the MASE follow the formulation in Section 2 of the paper? The current implementation seems utilize all historical data (instead of the data from 1~T), and the denominator in current implementation does not include future data, i.e. data from T~T+H?
In the MASE, I find that the shapes of
in_sample
,out_sample
, andforecasting
should betime_o
ortime_in
, instead ofbatch, time_i/o
. Moreover, does the MASE follow the formulation in Section 2 of the paper? The current implementation seems utilize all historical data (instead of the data from1~T
), and the denominator in current implementation does not include future data, i.e. data fromT~T+H
?