GenMRP
收藏资源简介:
GenMRP 是一个用于高效和个性化实时工业导航的生成式多路径规划框架的数据集。该数据集包含多种特征类型,用于支持表格分类和图机器学习任务。具体特征包括: 1. 链接特征(Link Features):包含路段属性,如路段长度和车道宽度,形状为 K * 2 * N。 2. 频率特征(Frequency Features):记录用户过去三个月的旅行历史,形状为 K * 2 * 10 * 7,包括时间差和历史状态等关键特征。 3. 场景特征(Scene Features):表示路径推荐的上下文信息,形状为 1 * 10,包括请求时间和用户对起点和终点的熟悉程度。 4. 长期特征(Long Features):按时间顺序排列的一系列路径选择记录,形状为 T * 31,包括选定和未选定路径的特征。该数据集适用于工业导航中的路径规划和个性化推荐任务。
GenMRP is a dataset for a generative multi-path planning framework enabling efficient and personalized real-time industrial navigation. This dataset includes multiple feature types to support tabular classification and graph machine learning tasks. The specific features are listed below: 1. Link Features: Contain road segment attributes such as segment length and lane width, with a shape of K * 2 * N. 2. Frequency Features: Record the user's travel history over the past three months, with a shape of K * 2 * 10 * 7, covering key features including time differences and historical states. 3. Scene Features: Represent the contextual information for route recommendations, with a shape of 1 * 10, including the request time and the user's familiarity with the origin and destination. 4. Long Features: A series of time-ordered route selection records, with a shape of T * 31, covering features of both selected and unselected routes. This dataset is suitable for route planning and personalized recommendation tasks in industrial navigation.




