International Journal of Research in Engineering and Innovation Vol-3, Issue-2 (2019), 79-81 __________________________________________________________________________________________________________________________________ International Journal of Research in Engineering and Innovation (IJREI) journal home page: http://www.ijrei.com ISSN (Online): 2456-6934 _______________________________________________________________________________________ Corresponding author: Hind khudheyer Email Address: hindkhab@gmail.com 79 http://doi.org/10.36037/IJREI.2019.3201 A new method combining discrete wavelet transform and neural network for high energy physics problem Hind Khudheyer 1 , Hisham Jashami 2 1 University of Karbala, Department of physics, Karbala, Iraq 2 University of Karbala, Department of Civil Engineering, Karbala, Iraq _______________________________________________________________________________________ Abstract In this paper, a new method proposed for high energy physics problem combining discrete wavelet transform and neural network. DWT used to decrease the size of features and the output of DWT wired to the neural network. The neural network applied to classify the features that obtained by DWT. Several statistical parameters are calculated to evaluate the performance of the proposed method. The proposed method presented remarkable results when compared with previous studies. © 2019 ijrei.com. All rights reserved Keywords: SVM, High-Energy Physics, Power of Signal _________________________________________________________________________________________________________ 1. Introduction High energy physics data examination is an honest multivariate problem. Notwithstanding the detail that multivariate methods have remained applied in High Energy Physics for an extended period, the volatile development of machine learning (ML) methods throughout the previous two decades had lone an incomplete influence on the familiar chic in which data investigation is achieved in this arena. Lone lately an augmented attention in the ML knowledge learned in other ranges of discipline can be experiential. ML applied to different fields successfully such as face recognition, image classification, and data classification [1]. Then, in recent years’ number of studies presented which applied ML to high energy physics. Because the high energy physics is difficult and hot topic exactly in the last years, then, we developed software has ability to classify the high energy physics problem in low computation time and high accurate. 2. Material & Methods 2.1 Discrete Wavelet Transform (DWT) The discrete wavelet transform (DWT) is an application of the wavelet transform (WT) using a discrete set of the wavelet scales and conversions submitting some definite rules. On other hand, this transmute decays the signal into equally orthogonal set of wavelets, which is the chief variance from the continuous wavelet transform (CWT), or its application for the discrete time series occasionally called discrete-time continuous wavelet transform (DT-CWT) [2][1]. The process of DWT presented in Figure 1. (, ) = 1 √ 0  =∑ () ( ( −  0    0 . )  0  ) (1) ∞ =−∞ Where j, k, nZ and a0 > 1 2.2 Neural Network ANN is a data treating model that is stimulated by the way biological nervous organizations, like the brain, procedure data. The key component of this model is the original organization of the data treating organization. It is collected of a great amount of extremely organized processing rudiments employed in agreement to resolve exact difficulties. ANN, like individuals, learn by instance.