Dynamic Prediction Length for Time Series with Sequence to Sequence Networks
Machine Learning
2019-08-20 v2 Artificial Intelligence
Machine Learning
Abstract
Recurrent neural networks and sequence to sequence models require a predetermined length for prediction output length. Our model addresses this by allowing the network to predict a variable length output in inference. A new loss function with a tailored gradient computation is developed that trades off prediction accuracy and output length. The model utilizes a function to determine whether a particular output at a time should be evaluated or not given a predetermined threshold. We evaluate the model on the problem of predicting the prices of securities. We find that the model makes longer predictions for more stable securities and it naturally balances prediction accuracy and length.
Cite
@article{arxiv.1807.00425,
title = {Dynamic Prediction Length for Time Series with Sequence to Sequence Networks},
author = {Mark Harmon and Diego Klabjan},
journal= {arXiv preprint arXiv:1807.00425},
year = {2019}
}