> ML_LITERATURE // NIE-2023-TIME-SERIES-WORTH-64-WORDS-PATCHTST_v1.0
A Time Series is Worth 64 Words: Long-term Forecasting with Transformers (PatchTST)
Yuqi Nie, Nam H. Nguyen, Phanwadee Sinthong, Jayant Kalagnanam · International Conference on Learning Representations (ICLR) (2023)
seminal-architecture2023industry-standardthirdPartyReproduced
Principal Contribution
Segmented sub-series into patches fed as input tokens to a channel-independent transformer, drastically reducing memory complexity and capturing local temporal correlations.
Operational Relevance
Serves as qualified reference for implementing task-time-series-forecasting in production systems.
Assumptions
- Underlying spatio-temporal continuity and domain distributional stability hold
Limitations
- Performance scaling and computational footprint depend on receptive field depth, sequence length, and resolution
Connected Algorithms, Architectures & Tools
Related Algorithms:
Related Architectures:
Implementing Libraries: