We lifted 11 functions out of this paper's own repositories and ran 7 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.
| Repository | Role | Ran |
|---|---|---|
| salesforce/CoST | canonical | 7 of 11 |
| Function | Status | Where it lives |
|---|---|---|
| cal_metrics | Ran | salesforce/CoST/tasks/forecasting.py code served (permissive licence) · get_code("a4750195b8805692") |
| fit_ridge | Ran | salesforce/CoST/tasks/_eval_protocols.py code served (permissive licence) · get_code("1d3809217bd3608f") |
| generate_binomial_mask | Ran | salesforce/CoST/models/encoder.py code served (permissive licence) · get_code("fd10054ec4772b80") |
| generate_continuous_mask | Ran | salesforce/CoST/models/encoder.py code served (permissive licence) · get_code("acda587a1d1be3bf") |
| pad_nan_to_target | Ran | salesforce/CoST/utils.py code served (permissive licence) · get_code("5254eb44ce12370e") |
| pkl_load | Ran | salesforce/CoST/utils.py code served (permissive licence) · get_code("645d1fdb1dc0e80c") |
| torch_pad_nan | Ran | salesforce/CoST/utils.py code served (permissive licence) · get_code("79783deb10f33397") |
| eval_forecasting | Not yet run | salesforce/CoST/tasks/forecasting.py code served (permissive licence) · get_code("a0b497709017e0db") |
| generate_pred_samples | Not yet run | salesforce/CoST/tasks/forecasting.py code served (permissive licence) · get_code("63e193a1a856ae88") |
| load_forecast_csv | Not yet run | salesforce/CoST/datautils.py code served (permissive licence) · get_code("3dc27cba7950c088") |
| load_forecast_npy | Not yet run | salesforce/CoST/datautils.py code served (permissive licence) · get_code("ee0abcf391ea662f") |
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
Deep learning has been actively studied for time series forecasting, and the mainstream paradigm is based on the end-to-end training of neural network architectures, ranging from classical LSTM/RNNs to more recent TCNs and Transformers. Motivated by the recent success of representation learning in computer vision and natural language processing, we argue that a more promising paradigm for time series forecasting, is to first learn disentangled feature representations, followed by a simple regression fine-tuning step -- we justify such a paradigm from a causal perspective. Following this principle, we propose a new time series representation learning framework for time series forecasting named CoST, which applies contrastive learning methods to learn disentangled seasonal-trend representations. CoST comprises both time domain and frequency domain contrastive losses to learn discriminative trend and seasonal representations, respectively. Extensive experiments on real-world datasets show that CoST consistently outperforms the state-of-the-art methods by a considerable margin, achieving a 21.3% improvement in MSE on multivariate benchmarks. It is also robust to various choices of backbone encoders, as well as downstream regressors. Code is available at https://github.com/salesforce/CoST.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2202.01575")
get_code_for_paper("2202.01575")
have("2202.01575")
Connect an agent — have() is free.