四个开源 Granite 模型、一个可扩展的预测接口,以及更少的模型选择不确定性。预训练时间序列模型无需在特定数据集上训练即可进行预测,但仍需从中选择一个模型。擅长处理平滑工业信号的模型可能在应对噪声较大的零售需求或长季节性周期时表现不佳,最佳选择取决于具体的时间序列和任务。
Four open Granite models, one extensible forecasting interface, and less guesswork about which model to use. Pre-trained time-series models can forecast without training on your dataset, but you still have to choose one. A model that handles smooth industrial signals well may struggle with noisy retail demand or a long seasonal horizon, and the best choice depends on the series and the task.
推出四个开源 Granite 模型、一个可扩展的预测接口,并减少关于模型选择的猜测。预训练时间序列模型无需在您的数据集上进行训练即可进行预测,但您仍需选择一个模型。擅长处理平滑工业信号的模型可能在应对噪声较大的零售需求或长季节性周期时表现不佳,最佳选择取决于具体的时间序列和任务。`result.prediction_outputs` 包含点预测值,而 `result.quantile_outputs` 包含分位数预测值,默认范围为 0.1 至 0.9。应用程序可以使用点值作为中心预测,并利用选定的分位数来表示预测范围。例如,0.1 和 0.9 分位数定义了一个名义覆盖率为 80% 的中心预测区间。这些架构以不同方式捕捉趋势、季节性和局部变异,因此它们的误差未必重合。一项涵盖 50 年的预测组合综述描述了如何通过组合预测来降低选错单一模型的风险,同时强调成员质量和多样性仍然至关重要。每个成员返回分位数预测。两种主要配方使用不同的规则将这些预测转化为一个预测分布及其均值。概率集成(均匀概率池化)赋予每个成员同等影响力,在概率空间中对它们的分位数预测进行池化,并从合并后的值中提取所需的分位数。由于权重是固定的,因此没有拟合步骤,也无需为新序列调整加权超参数。预测组合难题提供了有用的动机:在点预测中,简单的等权组合往往能很好地抵御那些需要估计权重的方法。该配方将同样的等影响力理念应用于概率预测。它是上述示例中加载的默认配方。较窄的区间并不一定意味着更准确或校准更好的预测,且 IQR 加权可能会偏向过度自信的成员。如果您希望进行实验,该配方暴露了温度设置(较高值使权重趋向均匀)以及对任何单个成员权重的可选上限。建议从概率集成开始。如果预测离散度对您的应用具有信息价值,请在代表性数据上将 IQR 集成与其进行比较,因为准确性和校准取决于序列和预测视野。排名列计算的是相对于完整 GIFT-Eval 参考排行榜的平均每配置排名,而不仅仅是这六个预测,因此随着排行榜条目的增加,排名值可能会发生变化。该仓库还包含了添加 IBM Research 检查点的扩展配方。这些检查点带有单独的许可条款,包括非商业和研究使用限制,因此在使用前请审阅每个成员的模型卡。GIFT-Eval 是一个广泛的基准测试,但总体增益并不意味着每个序列都能受益,平均排名也不能确立校准效果。了解配方在您应用中行为方式的途径是在代表性数据上对其进行回测:加载默认配方,以便在一次调用中获取来自所有四个 Granite 架构的中位数和分位数预测,并根据您的需求演变调整配方。
Four open Granite models, one extensible forecasting interface, and less guesswork about which model to use. Pre-trained time-series models can forecast without training on your dataset, but you still have to choose one. A model that handles smooth industrial signals well may struggle with noisy retail demand or a long seasonal horizon, and the best choice depends on the series and the task. result.prediction_outputs contains the point forecast, while result.quantile_outputs contains the quantile predictions which by default range from 0.1 to 0.9. Applications can use the point value as a central prediction and selected quantiles to express forecast ranges. For example, the 0.1 and 0.9 quantiles define a central forecast interval with nominal 80% coverage. These architectures capture trends, seasonality, and local variation in different ways, so their errors need not coincide. A 50-year review of forecast combinations describes how combining forecasts can reduce the risk of picking the wrong single model, while stressing that member quality and diversity still matter. Each member returns quantile forecasts. The two main recipes use different rules to turn those forecasts into one predictive distribution and its median. The Probabilistic Ensemble (uniform probability pooling) gives each member equal influence, pools their quantile forecasts in probability space, and extracts the requested quantiles from the combined values. Because the weights are fixed, there is no fitting step and no weighting hyperparameters to tune for a new series. The forecast combination puzzle offers useful motivation: in point forecasting, simple equal-weight combinations have often held up well against methods with estimated weights. This recipe applies the same equal-influence idea to probabilistic forecasts. It is the default loaded in the example above. A narrower interval does not necessarily mean a more accurate or better-calibrated forecast, and IQR weighting can favor an overconfident member. If you want to experiment, the recipe exposes a temperature setting (higher values move the weights toward uniform) and an optional cap on any single member's weight. Start with the Probabilistic Ensemble. If forecast spread looks informative for your application, compare the IQR Ensemble against it on representative data, since accuracy and calibration depend on the series and the horizon. Rank columns average per-configuration ranks against the full GIFT-Eval reference leaderboard, not just these six forecasts, so rank values can shift as entries are added to the leaderboard. The repository also includes expanded recipes that add IBM Research checkpoints. These carry separate license terms, including non-commercial and research-use restrictions, so review each member's model card before using them. GIFT-Eval is a broad benchmark, but aggregate gains don’t mean every series benefits, and average ranks don’t establish calibration. The way to find out how the recipes behave for your application is to backtest them on representative data: Load the default recipe to get median and quantile forecasts from all four Granite architectures in one call, and adapt the recipe as your needs evolve.