This message was deleted.
# neural-forecast
s
This message was deleted.
c
Hi @Phil! You understanding on both parameters is correct. Regarding the inverse, not necessarily. The
n_pool_kernel_size
downsamples only the inputs, and the
n_freq_downsample
only the output. Having both at the same time do not compound, because they affect different parts of the architecture. Here is the diagram of the paper, the kernel controls the Maxpool im the inputs of the MLP stack, and
n_freq_downsample
controls the output dimension of theta (points in the forecasting window)
The intuition of having both larger at the same time is that to output a lower dimensional output (higher
n_freq
) you need less information from the inputs, so kernel is larger
With that said, we have observed that larger kernel sizes only help in very high frequency data, and usually keeping a value of 1 or 2 is the best. That is why in our default config we kept the option of no downsampling ([1,1,1])
p
I see that makes more sense! Thank you! Have you observed any effects of varying the number of blocks across frequencies. or keeping them constant is roughly equivalent in performance
c
We recommend increasing the blocks with larger datasets. For example, the NBEATS uses 30 blocks in total for each frequency of the M4 dataset, with around 30k series
1