Temporal Fusion Transformers for Interpretable Multi-horizon Time Series Forecasting · HackerPapers discussions

Temporal Fusion Transformers for Interpretable Multi-horizon Time Series Forecasting — mentioned by gardnr

Mon, 16 Sep 2024 19:34:03 GMT

I was surprised: https://arxiv.org/abs/1912.09363 It's not exactly an LLM but it works in a similar fashion.

Paper and discussion history. A mention is not an endorsement.

Temporal Fusion Transformers for Interpretable Multi-horizon Time Series Forecasting — mentioned by spmurrayzzz

Wed, 17 Jul 2024 16:54:33 GMT

I can't speak for all use cases, but I've done a great deal of work in the space of using deep learning approaches for anomaly detection in network device telemetry. In particular with high resolution univariate time series of latency measurements, we saw success using convolutional autoencoders and GANs. These methods lean on reconstruction loss rather than forecasting, but still effective. There is some prior art for this that we leaned on [1][2]. RE: transformers — I did some early experimentation with Temporal Fusion Transformers [3] which worked pretty well for forecasting compared to…

Paper and discussion history. A mention is not an endorsement.

Temporal Fusion Transformers for Interpretable Multi-horizon Time Series Forecasting — mentioned by rdli

Wed, 16 Aug 2023 16:57:35 GMT

Temporal Fusion Transformer: see https://arxiv.org/abs/1912.09363.

Paper and discussion history. A mention is not an endorsement.

Temporal Fusion Transformers for Interpretable Multi-horizon Time Series Forecasting — mentioned by isoprophlex

Thu, 01 Jul 2021 06:30:11 GMT

Temporal Fusion Transformers look very cool: accepts timeseries vectors but also categoricals and numerical features, outputs distribution using quintile regression. https://arxiv.org/pdf/1912.09363v2.pdf Available as pytorch implementation: https://pytorch-forecasting.readthedocs.io/en/latest/index.h...

Paper and discussion history. A mention is not an endorsement.