An Anomaly Detection and Analysis Method for Network Traffic Based on Correlation Coefficient Matrix

2009 ◽  
Author(s):  
Ning Chen ◽  
Xiao-Su Chen ◽  
Bing Xiong ◽  
Hong-Wei Lu
Keyword(s):  
Anomaly Detection ◽  
Correlation Coefficient ◽  
Network Traffic ◽  
Coefficient Matrix ◽  
Analysis Method ◽  
Correlation Coefficient Matrix

Choose the Best Coach Based on the Theory of Grey System

2014 ◽  
Vol 998-999 ◽  
pp. 1757-1760 ◽  
Author(s):  
Jia Zhao
Keyword(s):  
Correlation Coefficient ◽  
Evaluation Model ◽  
Relative Size ◽  
Coefficient Matrix ◽  
Weight Matrix ◽  
Target Range ◽  
Grey System ◽  
Average Value ◽  
Dynamic Weight ◽  
Correlation Coefficient Matrix

In order to select “the best all time college coach” across both genders and all possible sports, we establish a multi-hierarchy evaluation model based on the theory of grey system. Without affecting the reliability of the results, we analyze the distribution of every index to determine the initial screening standard, narrow target range and simplify the problem. Then, we use the theory of grey system to evaluate coaches. We first normalize all indexes to make them comparable. Second, we determine a set of the best indexes as the reference vector and obtain the correlation coefficient matrix. On the issue of the weight of all indexes, we introduce the concept of dynamic weight matrix by considering the distribution and relative size of them. After that, we obtain the value of the final evaluation from the correlation coefficient matrix and dynamic weight matrix .As for the influence of time on the evaluation, we regroup the data of coaches according to the time section for five years, which results in some statistical variables, and then we draw the trend graph of them. To sum up, the number of excellent coaches of every sport is increasing over time, but the average value and standard deviation of winning percentage is decreasing. The model has the adaptability on different sports by comparison. And because the data of women coaches is insufficient, the influence of genders on the evaluation needs to be researched further.

Trend Extraction in Vibration Signal Based on EMD

2012 ◽  
Vol 459 ◽  
pp. 377-380
Author(s):  
Yu Hua Dong ◽  
Jun Xing Zhang
Keyword(s):  
Correlation Coefficient ◽  
Empirical Mode Decomposition ◽  
Vibration Signal ◽  
Coefficient Matrix ◽  
Trend Extraction ◽  
Mode Decomposition ◽  
Effective Combination ◽  
Correlation Coefficient Matrix

This paper proposed a de-trend method for vibration signal of telemetry based on the empirical mode decomposition (EMD) by correlation coefficient matrix. The signal is decomposed to a series of intrinsic mode component and the remainder item by EMD. It mainly distinguishes between the remainder item and the signal trend, according to the correlation coefficient matrix to determine whether some intrinsic mode component belongs to the trend item or not. The results show that signal trends can be extracted accurately through the effective combination EMD with correlation coefficient matrix and the proposed method has good applicability for different signals and different trends.

Correlation Coefficient Matrix Based PD Feature Dimensional Reduction

2014 ◽  
Vol 945-949 ◽  
pp. 2499-2504
Author(s):  
Yu Wang ◽  
Jin Sha Yuan ◽  
Hai Kun Shang ◽  
Song Jin
Keyword(s):  
Correlation Coefficient ◽  
Power Transformer ◽  
Partial Discharge ◽  
Back Propagation ◽  
Coefficient Matrix ◽  
Back Propagation Neural Network ◽  
Small Samples ◽  
Diagnostic Feature ◽  
Dimension Characteristic ◽  
Correlation Coefficient Matrix

The abstraction of diagnostic feature from field condition monitoring data is a significant research challenge. A new dimension reduction method based on correlation coefficient matrix is proposed aimed at the high-dimension characteristic parameters in the process of pattern recognition for partial discharge in power transformer. The CCM is constructed by parameters extracted from partial discharge signature in power transformer. The parameters that have similar classification characters are reduced directed by the correlation analysis result. The reduced PD features are inputted to the pattern classifiers of probabilistic neural networks (PNN). The results show that the parameter dimension is reduced and the classifier construction is simplified, and the recognition effect is better than that of the traditional back propagation neural network (BPNN) in the condition of small samples.

A New Method of Porous Space Reconstruction Using Multipoint Histogram Technology

2017 ◽  
Vol 140 (3) ◽  
Author(s):  
Na Zhang ◽  
Qian Sun ◽  
Mohamed Fadlelmula ◽  
Aziz Rahman ◽  
Yuhe Wang
Keyword(s):  
Correlation Coefficient ◽  
Pore Space ◽  
Coefficient Matrix ◽  
Training Image ◽  
Porous Space ◽  
Stable Theory ◽  
Pore Scale ◽  
New Construction ◽  
Point Correlation ◽  
Correlation Coefficient Matrix

Pore-scale modeling is becoming a hot topic in overall reservoir characterization process. It is an important approach for revealing the flow behaviors in porous media and exploring unknown flow patterns at pore scale. Over the past few decades, many reconstruction methods have been proposed, and among them the simulated annealing method (SAM) is extensively tested and easier to program. However, SAM is usually based on the two-point probability function or linear-path function, which fails to capture much more information on the multipoint connectivity of various shapes. For this reason, a new reconstruction method is proposed to reproduce the characteristics of a two-dimensional (2D) thin section based on the multipoint histogram. First, the two-point correlation coefficient matrix will be introduced to determine an optimal unit configuration of a multipoint histogram. Second, five different types of seven-point unit configurations will be used to test the unit configuration selection algorithm. Third, the multipoint histogram technology is used for generating the porous space reconstruction based on the prior unit configuration with a different calculation of the objective function. Finally, the spatial connectivity, patterns reproduction, the local percolation theory (LPT), and hydraulic connectivity are used to compare with those of the reference models. The results show that the multipoint histogram technology can produce better multipoint connectivity information than SAM. The reconstructed system matches the training image very well, which reveals that the reconstruction captures the geometry and topology information of the training image, for instance, the shape and distribution of pore space. The seven-point unit configuration is enough to get the spatial characters of the training image. The quality of pattern reproduction of the reconstruction is assessed by computing the multipoint histogram, and the similarity is around 97.3%. Based on the LPT analysis, the multipoint histogram can describe the anticipated patterns of geological heterogeneities and reproduce the connectivity of pore media with a high degree of accuracy. The two-point correlation coefficient matrix and a new construction theory are proposed. The new construction theory provides a stable theory and technology guidance for the study of pore space development and multiphase fluid flow rule in the digital rock.

scholarly journals Seasonal Distribution of Phytoplankton in Riwada Reservoir, Visakhapatnam, Andhra Pradesh, India

2013 ◽  
Vol 5 (3) ◽  
pp. 290-295 ◽  
Author(s):  
Jyothi KAPARAPU ◽  
Mohan Narasimha Rao GEDDADA
Keyword(s):  
Correlation Coefficient ◽  
Andhra Pradesh ◽  
Negative Relationship ◽  
Oxygen Demand ◽  
Diversity Indices ◽  
Coefficient Matrix ◽  
Total Hardness ◽  
Species Evenness ◽  
Total Phytoplankton ◽  
Correlation Coefficient Matrix

The present study deals with seasonal variations, correlation coefficient and biodiversity indices of phytoplankton during April 2011 to March 2012 in the Riwada reservoir, Visakhapatnam, Andhra Pradesh, India. Sampling was performed at five stations during pre-monsoon, monsoon and post monsoon. There were a total of 57 genera belonging to four major groups i.e., Chlorophyceae (27 genera), Bacillariophyceae (14 genera), Cyanophyceae (13 genera) and Euglenophyceae (three genera). Maximum and minimum total phytoplankton population and percentages were recorded at station three in pre monsoon and at station two during monsoon. The maximum and minimum species richness (Menhinick index R2) were found to be 1.29 at station one and 1.10 at station three respectively. Maximum and minimum species diversity (H1) were found at station four (3.98) and station two (3.71). Maximum species evenness was recorded at stations one, being four and five; minimum species evenness was recorded at station two. Correlation coefficient matrix indicated significant positive relationship with water temperature, pH, transparency, biological oxygen demand and chlorides, negative relationship with electric conductivity, total solids, total dissolved solids, total hardness, dissolved oxygen, nitrates, sulphates and phosphates of water. The diversity indices showed that the reservoir have a well balanced phytoplankton community.

scholarly journals A Character of Rotating Machinery Defined Based on KICA

2011 ◽  
Vol 2-3 ◽  
pp. 761-764
Author(s):  
Ling Li Jiang ◽  
Zong Qun Deng ◽  
Si Wen Tang
Keyword(s):  
Correlation Coefficient ◽  
Failure Modes ◽  
Rolling Bearing ◽  
Coefficient Matrix ◽  
Rotating Machinery ◽  
Good Effect ◽  
Gear Fault ◽  
Sensitive Character ◽  
Machinery Fault Diagnosis ◽  
Correlation Coefficient Matrix

This research aims at defining a character of rotating machinery——KIC, that the typical rolling bearing and gear failure modes can be effectively identified by using the character. Firstly, A correlation coefficient matrix is composed by the correlation coefficient between two-two inde-pendent component derived from kernel independent component (KICA). Then the KIC is defined by the correlation coefficient matrix. Experimental results show that the KIC has a good effect for identifying the bearing and gear fault modes, so it can be used as sensitive character for rotating machinery fault diagnosis.

Robust Principal Component Analysis Algorithm Based on Nonparametric Correlation Coefficient Matrix

2014 ◽  
Vol 989-994 ◽  
pp. 2613-2616
Author(s):  
Nan Zhao ◽  
Hong Yu Shao
Keyword(s):  
Principal Component Analysis ◽  
Correlation Coefficient ◽  
Numerical Experiments ◽  
Principal Component ◽  
Component Analysis ◽  
Coefficient Matrix ◽  
Analysis Algorithm ◽  
Principal Component Analysis Algorithm ◽  
Nonparametric Correlation ◽  
Correlation Coefficient Matrix

Principal component analysis algorithm is widely applied in solving a large number of dimension reduction problems for high-dimensional data in fields such as image identification and comprehensive evaluation due to its clear conception and simple application. The PCA based on covariance matrix and correlation coefficient matrix is very sensitive to abnormal values so that the data structure is distorted. As a result, there exist great deviations in the analysis results and it is hard to explain. The traditional PCA is remade in this paper based on non-parametric statistical ideas, nonparametric correlation coefficient and gray relational coefficient, in order to define three kinds of robustness evaluation criteria and objectively evaluate the robustness of the algorithm through numerical experiments. The results of the numerical experiments show that the new algorithm is of good robustness to noise pollution and the algorithm is of good applicability.

Sign in / Sign up

Export Citation Format

Share Document