The Research Of High-energy Physics Particle Classification Using The Neural Network Method

High-energy physics is a branch of physics disciplines. This work is on the structure and properties of microscopic world, which is deeper than the nuclei, as well as the phenomenon that these substances into each other at very high energies. And exploring the reason for these phenomena Oand laws are also included. Based on experimental high-energy physics-based disciplines. High Energy Physics experiments generally need to record and process large amounts of data which is used to describe the system characteristics and operating status. These descriptions portray a number of indicator variables of system features and with a huge number of samples, and with a certain degree of randomness, which make the sample data formed a huge and vastly complex data ocean. For complex data processing, in recent years, the neural network method is used in the field of high energy physics.There are two aspects of neural network applications in high energy physics. One is the triggering online and the other is the offline analysis. In the former, neural network can be integrated in processing chip in the detector, and can greatly improve sentenced rate benefiting from its large number of small parallel processing units. These chips have been used in many high-energy physics experiments. For the later one, neural network is widely used in data processing due to its unique ability of parallel computing. In this paper, we discuss using the three-layered BP neural network to identify the quark-gluon jets. The experiments indicate that the neural network can not only effectively identify the quark and gluon jets, but also ensure grate purity. The results show that neural network has grate validity and applicability in the identification of quark-gluon jets.At the same time, the neural network is applied to the distinction between signal and background. Compared with the result of the Fisher Linear Discriminant method, the results of the neural network have a largely advantage which shows that the neural network is efficiency even for the grate overlap samples.The last part describes the application of bagging algorithm in the neural network to analysis the experimental data in the high-energy physics. We still use the quark-gluon jets data for sample. And the results of the two treatments were compared. It shows that the use of bagging algorithm has greatly improved the sentence separation efficiency quark-gluon jets, including the identification efficiency, as well as purity and signal-to-noise ratio.