Shulou(Shulou.com)05/31 Report--

In this issue, the editor will bring you about how to analyze the data anti-counterfeiting technology based on SVM and browser features. The article is rich in content and analyzed and described from a professional point of view. I hope you can get something after reading this article.

Preface

In the field of security attack and defense, attackers often simulate normal business requests by falsifying browser information (such as userAgent, cookie, etc.), so as to automatically simulate repeated requests, thus creating violent access, information traversal, and wool retrieval. This kind of data forgery has bad influence and is difficult to identify. The following is an effective way to intelligently identify data forgery by integrating machine learning algorithm with browser gene technology.

Support vector machine (SupportVector Machine,SVM)

Support vector machine, namely SVM method (Support Vector Machine), was proposed by Vapnik et al in 1995. It has relatively good performance index. This method is a machine learning method based on statistical learning theory. Through the learning algorithm, SVM can automatically find the support vectors that have better ability to distinguish the classification, and the classifier constructed from this classifier can maximize the interval between classes, so it has better adaptability and higher accuracy. This method only needs to determine the final classification result by the category of boundary samples of all kinds of domains. The purpose of the support vector machine algorithm is to find a hyperplane H (d), which can separate the data in the training set and has the largest distance from the boundary of the class domain along the direction perpendicular to the hyperplane, so the SVM method is also called the maximum edge (maximum margin) algorithm. Most of the samples in the sample set to be separated are not support vectors, and removing or reducing these samples has no effect on the classification results. SVM method has better classification results for automatic classification in the case of small samples.

The SVM method uses a non-linear mapping p to map the sample space to a high-dimensional or even infinite-dimensional feature space (Hilbert space), so that the problem of nonlinear separability in the original sample space is transformed into a linearly separable problem in the feature space. To put it simply, it is promotion and linearization. Dimensionality enhancement is to map samples to high-dimensional space, which will generally increase the complexity of computation and even cause "dimension disaster", so people seldom pay attention to it. However, as for the problems of classification and regression, it is likely that the sample set which can not be linearly processed in the low-dimensional sample space can be linearly divided (or regressed) by a linear hyperplane in the high-dimensional feature space. General dimensionality enhancement will bring computational complexity, and SVM method skillfully solves this problem: by applying the expansion theorem of kernel function, there is no need to know the explicit expression of nonlinear mapping; because the linear learning machine is established in high-dimensional feature space, compared with the linear model, it not only does not increase the computational complexity, but also avoids the "dimension disaster" to some extent. All this is due to the expansion of kernel functions and the theory of calculation.

The two most important points in SVM are kernel function and solving algorithm SequentialMinimal Optimization (SMO for short).

Kernel function

1. Method principle

According to the pattern recognition theory, linear inseparable patterns in low-dimensional space may be linearly separable by nonlinear mapping to high-dimensional feature space, but if this technique is directly used for classification or regression in high-dimensional space, there are some problems, such as determining the form and parameters of the nonlinear mapping function and the dimension of the feature space, and the biggest obstacle is the "dimension disaster" in the operation of the high-dimensional feature space. This problem can be solved effectively by using kernel function technology.

Let X ∈ X belong to R (n) space, and the nonlinear function Φ realizes the mapping from X to F between inputs, where F belongs to R (m), n.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.