infinite-dataset-hub/InventoryAnomalies

InventoryAnomalies数据集是一个精心策划的零售公司库存记录集合，涵盖了多年的数据。每条记录代表一个库存事件，如库存水平更新，包含项目ID、项目名称、更新日期、数量和位置等详细信息。数据集还包括一个label列，指示库存水平变化是否为异常（即异常值或需求激增）或正常。标签是基于历史趋势、季节性和已知需求激增生成的。该数据集的目的是帮助机器学习从业者训练模型以进行库存管理中的异常检测。

The InventoryAnomalies dataset is a curated collection of inventory records from a retail company that spans several years. Each record represents an inventory event, such as a stock level update, which includes details like item ID, item name, date of the update, quantity, and location. The dataset also includes a label column indicating whether the stock level change is an anomaly (i.e., an outlier or demand spike) or normal. The labels are generated based on historical trends, seasonality, and known demand spikes. The aim is to assist ML practitioners in training models for anomaly detection in inventory management.