Recent progress in neural forecasting instigated significant improvements in the accuracy of large-scale forecasting systems. Yet, extremely long horizon forecasting remains a very difficult task. Two common challenges afflicting the long horizon forecasting are the volatility of the predictions and their computational complexity. In this paper we introduce N-HiTS, which addresses both challenges by incorporating novel hierarchical interpolation and multi-rate data sampling techniques. These techniques enable our method to assemble its predictions sequentially, selectively emphasizing components with different frequencies and scales while decomposing the input signal and synthesizing the forecast. We conduct an extensive empirical evaluation demonstrating the advantages of N-HiTS over the state-of-the-art long-horizon forecasting methods. On an array of multivariate forecasting tasks, our method provides an average accuracy improvement of 25% over the latest Transformer architectures while reducing the computational time by orders of magnitude.
N-HiTS architecture. The model is composed of several MLPs with ReLU nonlinearities. Blocks are connected via doubly residual stacking principle with the backcast y[t-L:t, l] and forecast y[t+1:t+H, l] outputs of the l-th block.
Multi-rate input pooling, hierarchical interpolation and backcast residual connections together induce the specialization of the additive predictions in different signal bands, reducing memory footprint and compute time, improving architecture parsimony and accuracy.
To replicate the results of the paper, in particular to produce the forecasts for N-HiTS, run the following:
make initmake get_datasetto download data.
make run_module module="python -m nhits_multivariate --hyperopt_max_evals 10 --experiment_id run_1"If you want to use GPU simply add gpu=0 to the last line.
make run_module module="python -m nhits_multivariate --hyperopt_max_evals 10 --experiment_id run_1" gpu=0- Evaluate results for a dataset using:
make run_module module="python -m evaluation --dataset ETTm2 --horizon -1 --model NHITS --experiment run_1"Alternatively, run all evaluations at once:
for dataset in ETTm2 ECL Exchange traffic weather ili;
do make run_module module="python -m evaluation --dataset $dataset --horizon -1 --model NHITS --experiment run_1";
done