Classification Method of Warehouse Inventory Data Based on PE-AdaBoost Algorithm

Abstract

In view of the problems existing in traditional warehouse inventory data mining, such as low classification efficiency, lack of dynamic adjustment ability of weak classifiers, and excessive weight of samples with multiple classification errors, a perceptual enhanced classification algorithm based on AdaBoost is proposed. First, the multi-layer perceptual network is used to dynamically adjust the weight of the weak classifier, complete the sample weighting in the first perceptual network, and complete the weight calculation of the weak classifier in the second perceptual network, and obtain the best weight. Then, the weight attenuation superparameter is introduced to process the sample weight, which is used to solve the phenomenon that the samples with multiple classification errors are set with too high weights. Through the comparison with other warehouse dataset classification methods, it is found that this method can improve the accuracy of the current classification methods, and is superior to other algorithms in terms of running performance.