---
license: apache-2.0
task_categories:
- time-series-forecasting  # 这是让系统识别为时间序列任务的关键
tags:                      # 自定义标签
- time-series
- forecasting
- benchmark
---

This repository contains the extracted time series features (tsfeatures) for each variate and the detailed forecasting results for every experiment.

**Note**: These files are for building leaderboard and visualization; users do not need to download this directory.

* `features/`: **Statistical Features (tsfeatures)**

    Each dataset's features are saved to: `output/features/{dataset}/{freq}/`.
    This directory stores the computed tsfeatures for the variates in the dataset. The folder contains a CSV file named either:
    * `test.csv`: Features calculated on the **test split**.
    * `full.csv`: Features calculated on the **full dataset**.

* `results/`: **Experimental Results**

    Each experiment is saved to: `output/results/{model}/{dataset}/{freq}/{horizon}/`.
    Each directory represents a specific model-dataset-frequency-horizon combination and contains the following files:
    * `config.json`: The experiment settings and hyperparameters.
    * `metrics.npz`: Window-level forecasting metrics (e.g., MASE, CRPS).
    * `predictions.npz`: Quantile prediction results for each sliding window.