The Combination Of Statistics And AI And Its Applications

In this paper,I use face recognition technology as a link to combine statistical ideas with Artificial Intelligence.First of all,the PCA method in statistics is integrated into the face recognition technology to make full use of the idea of dimension reduction,construct the feature face space,and ultimately achieve the purpose of efficient recognition.Then,the existing local Binary Pattern(LBP)method was introduced as a conventional comparison item.At the same time,on the basis of existing Kernel Principal Component Analysis(KPCA)and Support Vector Machine(SVM),the advantages of the two are complementary.That is to say,the signatures extracted by KPCA are used as input to the SVM,and the two classifications are performed.Forecasting fully embodies the idea of Cluster Analysis and further accelerates the retrieval speed of face recognition,and finally achieves the effect of 1+1>2.This is the key KPCA+SVM method proposed in this paper.Finally,the performance of PCA,LBP and KPCA+SVM are compared by simulation.Compared with the PCA method,the KPCA+SVM method will effectively improve the recognition accuracy and recognitio efficiency.However,compared with the LBP method,the recognition accuracy rate is still the key point for improvement.It is also the direction that I need to focus on in the futureThe whole system is implemented in MATLAB software and design a Graphical User Interface to complete the whole function of the system.