Method and system for label-free data classification and predication

Abstract

The invention discloses a method and system for label-free data classification and predication. History data is utilized and rigid requirements on history data labels are removed. Also, experience information database accumulated in a service process is combined and an aim of automatically improving predication precision can be achieved through subsequence continuous optimization of real data. Themethod includes inputting service flow data and acquiring service scene data in a service flow; inputting service content data and grouping the service content data according to the service scene data; constructing characteristic indexes of service content data according to the service scene data and a corresponding service content database; cleaning the characteristic indexes of the service content data; clustering the service content data subjected to characteristic index cleaning and determining the class centers of different classes; calculating the sampling weights of the different classes; sampling the service content data according to the sampling weights and marking predication labels on the sampling results.

Description

technical field [0001] The present invention relates to the technical field of data classification prediction, in particular to a method and system for unlabeled data classification prediction. Background technique [0002] Risk detection is a commonly used quality detection method. This method is widely used in business analysis in various industries to detect potential risks in business for early detection and control. For general enterprises or regulatory agencies, risk detection methods are mainly divided into three types: one is to use quality inspectors to inspect the tested objects one by one to discover the risks of the tested objects; The risk of the tested product; the third is to predict the probability of the risk of each tested object through the information data and historical data of the production of the product, and then conduct actual sampling inspections on the higher risk tested products. [0003] Among the three risk detection methods described above, t...

AI Technical Summary

Problems solved by technology

The usage scenario of the second detection method is similar to that of the first one. It is not suitable for products with many types of products and complex technologies. This method can count the proportion of qualified (normal) tested products, but it will let go of a certain proportion of risks. Detected product

Therefore, the first two detection methods mentioned above require a lot of manpower and time and are not suitable.

The third way to detect the risk of each tested product from the data requires historical label data, but due to various reasons, many systems have not stored the label data, so this method has the technical problem of low prediction accuracy, and, due to its It relies heavily on the labels that have been marked in historical data, and cannot be applied to the environment of unlabeled data prediction, nor can it be used in business scenarios for anomaly detection

Images