US Flight Data

该数据集包含了用于预测航班起飞是否延误的飞行记录，其中包括对近期及周围航班观察到的数据以及气象信息。此外，对8372次航班转移进行了评估，并以ROC曲线下的面积来衡量性能。数据规模涉及92个学习领域，包括10个机场和11家航空公司，每个领域有1,880到9,500个LSOs。该任务旨在预测航班起飞延误情况。This dataset contains flight records for flight takeoff delay prediction, including observed data from recent and adjacent flights as well as meteorological information. A total of 8,372 flight transfer events were evaluated for this task, with model performance measured using the Area Under the Receiver Operating Characteristic Curve (ROC-AUC). The dataset covers 92 learning domains, involving 10 airports and 11 airlines, where each domain contains between 1,880 and 9,500 LSOs. The core task of this dataset is to predict flight takeoff delays.

搜集汇总

背景与挑战

背景概述

该数据集是“US Flight Data”相关项目的一部分，主要用于知识驱动的转移学习解释研究，包含航班本体、代码示例和示例数据，但未直接提供原始航班数据，而是侧重于支持机器学习模型的转移学习和推理过程。