Depth anomaly detection method for label-free table data and electronic equipment

The invention discloses a depth anomaly detection method for label-free table data and electronic equipment, and the method comprises the following steps: constructing an abnormal data sample of a table, and combining the constructed abnormal data sample with a normal data sample of the table into a training set; converting the table structure data into graph structure data, and obtaining a corresponding adjacent matrix; taking the adjacent matrix as training data to train the depth map auto-encoder to obtain a trained depth map auto-encoder; and detecting a data sample of a target table by adopting the trained depth map auto-encoder to realize the depth anomaly detection of the label-free table data. The method has the advantages of high applicability, high recognition precision and full utilization of data information for different data sets. The problem that a traditional anomaly detection method is poor in adaptability to different data sets is solved, in actual use, personal errors caused by manually set