University, Samsun, Turkey b Statistics Department, Faculty of Science, Ankara University,

Size: px
Start display at page:

Download "University, Samsun, Turkey b Statistics Department, Faculty of Science, Ankara University,"

Transcription

1 This article was downloaded by: [Marmara Universitesi] On: 01 April 2015, At: 02:06 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: Registered office: Mortimer House, Mortimer Street, London W1T 3JH, UK Click for updates Communications in Statistics - Theory and Methods Publication details, including instructions for authors and subscription information: A New Robust Regression Method Based on Particle Swarm Optimization Ozge Cagcag a, Ufuk Yolcu b & Erol Egrioglu c a Statistics Department, Faculty of Arts and Science, Ondokuz Mayis University, Samsun, Turkey b Statistics Department, Faculty of Science, Ankara University, Ankara, Turkey c Statistics Department, Faculty of Arts and Science, Marmara University, Istanbul, Turkey Accepted author version posted online: 29 May To cite this article: Ozge Cagcag, Ufuk Yolcu & Erol Egrioglu (2015) A New Robust Regression Method Based on Particle Swarm Optimization, Communications in Statistics - Theory and Methods, 44:6, , DOI: / To link to this article: PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the Content ) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &

2 Conditions of access and use can be found at

3 Communications in Statistics Theory and Methods, 44: , 2015 Copyright Taylor & Francis Group, LLC ISSN: print / X online DOI: / A New Robust Regression Method Based on Particle Swarm Optimization OZGE CAGCAG, 1 UFUK YOLCU, 2 AND EROL EGRIOGLU 3 1 Statistics Department, Faculty of Arts and Science, Ondokuz Mayis University, Samsun, Turkey 2 Statistics Department, Faculty of Science, Ankara University, Ankara, Turkey 3 Statistics Department, Faculty of Arts and Science, Marmara University, Istanbul, Turkey Regression analysis is one of methods widely used in prediction problems. Although there are many methods used for parameter estimation in regression analysis, ordinary least squares (OLS) technique is the most commonly used one among them. However, this technique is highly sensitive to outlier observation. Therefore, in literature, robust techniques are suggested when data set includes outlier observation. Besides, in prediction a problem, using the techniques that reduce the effectiveness of outlier and using the median as a target function rather than an error mean will be more successful in modeling these kinds of data. In this study, a new parameter estimation method using the median of absolute rate obtained by division of the difference between observation values and predicted values by the observation value and based on particle swarm optimization was proposed. The performance of the proposed method was evaluated with a simulation study by comparing it with OLS and some other robust methods in the literature. Keywords Linear model; Particle swarm optimization; Robust regression estimator; Simulation. Mathematics Subject Classification 1. Introduction Due to its easy calculation, ordinary least squares (OLS) technique is the most common method used in regression analysis which is used for eliciting unknown values of dependent variable and modeling its behavior. OLS method which is used to estimate the unknown parameters of regression model focuses on minimizing difference between predicted and observation values. This makes OLS technique very sensitive to the outliers in data set and may cause deviations in the model. Received March 6, 2012; Accepted July 30, Address correspondence to Ufuk Yolcu, Statistics Department, Faculty of Arts and Science, Ankara University, 06100, Ankara, Turkey; uyolcu@ankara.edu.tr 1270

4 New Robust Regression Method Based on Particle Swarm Optimization 1271 In cases where data set has outliers, the detection of existing outlier observation and exclusion of the observation from the data set can be considered as a way of analyzing the model properly. However, as each observation in data set may affect the model significantly, exclusion of them from the data set will affect the result substantially and thus will change the model. Therefore, each observation in the data set should be used to reach the appropriate solution. For this purpose, in cases where data set has outliers, robust methods which are less sensitive than OLS technique are used. In literature, several studies have been carried out for this purpose. Techniques, known as M estimators, aim to minimize the function of residuals rather than the sum of the squares of the residuals. Hogg (1979), Huber (1981), Huynh (1982), Hampel et al. (1986), Dave and Krishnapuram (1997) proposed robust methods using M estimators. Besides, Rousseuw and Leroy (1997) and Candan (1995) proposed robust methods which make residual squares as median least squares. Akbilgic and Akinci (2009) proposed a new approach based on least squares ratio (LSR). Sanli (2005) proposed an approach that extends multiple regression analysis to simple linear regression analysis and tried to reduce the weights of outliers in model forecasting. In this study, a new parameter estimation method using median of absolute rate obtained by division of difference between observation values and predicted values by the observation value and based on particle swarm optimization (PSO) was proposed. The performance of the proposed method was evaluated with a simulation study by comparing it with OLS and some other robust methods in the literature. In the second section of the study, simple linear regression model and OLS technique is introduced. Third section deals with the proposed method. In the fourth section, implementation of the proposed method is introduced with a simulation study, and obtained results are discussed. 2. Multiple Linear Regression Model and OLS A linear regression model can be shown as expressed in (1). y i = β 0 + p x ji β j + e i i = 1, 2,...,n j = 1, 2,...,p (1) j=1 Here, y i represents dependent variable for the i th observation, x i1,x i2,,x ip represents the number independent variable value that equalsp, e i represents error, and β 0,β 1,β 2,,β p represents regression coefficients. The expression of this model with matrix notation can be expressed as: Y = Xβ + ε (2) where Y is an n 1 dimensional vector of dependent variable observation, X is an n (p+1) dimensional matrix of dependent variable observation, β; (p + 1) 1 dimensional vector of regression coefficient and ε is an n 1 dimensional error vector. OLS technique, used in forecasting model parameters (β), is based on minimization of Eq. (3). (Y Ŷ ) = ε ε = (Y Xβ) (Y Xβ) (3)

5 1272 Cagcag et al. In order to make ˆβ estimator blue (best linear unbiased) which was obtained from OLS technique, error should be 0 mean or σ 2 variance. Unbiased estimator of σ 2 is as in Eq. (4). ˆσ 2 = 1 n p (Y X ˆβ) (Y X ˆβ) (4) Residual analysis, widely used in detecting outlier in regression analysis, is an effective method. In the model, the residual of i th observation is calculated as: e i = y i ŷ i i = 1, 2,...,n (5) 3. Particle Swarm Optimization In this study, in cases where there is an outlier observation in data set, PSO algorithm is used for the purpose of parameter estimation. PSO technique was proposed by Kennedy and Eberhart (1995). PSO is a population-based optimization algorithm. The most important feature of this algorithm is its ability to reach the optimum point from the several points at the same time. Because of this feature, local optimum does not block it and it has a better chance to reach global optimum. Modified PSO method, whose algorithm given below, is used in the study. In the given algorithm, inertia weight was taken as time-varying as stated in Shi and Eberhart (1999). Similarly, time-varying acceleration coefficients were employed as in the study by Ma et al. (2006). Algorithm 1. Modified PSO Algorithm Step 1. Optimization of particles (xi k,i = 1, 2,...,d; k = 1, 2,...,pn) are selected randomly and stored in X. X = { x k 1,xk 2,...,xk d},k = 1, 2,...,pn (6) Here, p n represents the number of particle, whereas d shows the number of position in each particle. Step 2. Velocities are determined randomly and stored in V. V = { v1 k },vk 2,...,vk d (7) Step 3. pbest and gbest are generated based on performance function. pbest i = (p i1,p i2,...,p id ) i = 1, 2,...,d (8) pbest g = gbest = (p g1,p g2,...,p gd ) (9) Here, pbest are the best positions of particle in separate iterations whereas gbest are the best positions of all particles in a population. Step 4. Possible intervals are determined for, where w is inertia parameter, c 1 and c 2 are cognitive and social coefficients. Inertia parameter (w), cognitive (c 1 ) and social (c 2 ) coefficients are calculated in each iteration in accordance with the following formula. t c 1 = (c 1f c 1i ) maxt + c 1i (10)

6 New Robust Regression Method Based on Particle Swarm Optimization 1273 Figure 1. Presentation of a particle in PSO. t c 2 = (c 2f c 2i ) maxt + c 2i (11) w = (w 2 w 1 ) maxt t + w 1 maxt (12) Here, (c 1i,c 1f ) are possible intervals for cognitive coefficient, (c 2i,c 2f )forsocial coefficient, and (w 1,w 2 ) for inertia parameters, respectively. maxt represents maximal number of iteration, and t represents valid iteration number. Step 5. New velocities and positions are calculated according to the formula given below. v k+1 id = [ w vid k + c 1 rand 1 (pbest id x id ) + c 2 rand 2 (pbest gd x id ) ] x k+1 id = x id + v k+1 id where rand 1 and rand 2 are random numbers between 0 and 1. Step 6. Steps 1 5 are repeated until a predetermined number of iteration (maxt) is reached. 4. The Proposed Method OLS method, widely used in regression analysis for parameter estimation, is highly affected by the presence of outlier observations in data set. In literature, parameter estimations are calculated by giving outliers lower weights in M-estimators. Using a median rather than error mean may decrease the effect of outliers without need for weight values. In literature, the mean absolute error (y i ŷ i ) is minimized for parameter estimation by using least median squares approach. Additionally, when compared with the absolute error, proportional error is considered to be a more valid error criterion. Akbilgic and Akinci (2009) proposed a new approach that minimizes mean proportional error. In this study, a method based on PSO which cannot be affected by outliers in data set in forecasting model parameter was proposed. The proposed method relies on minimizing the median of proportional error differently from M-estimator, least median squares and Akbilgic and Akinci s method. The proposed method uses the expression given in (15) as an objective function. ( min median β y i ŷ i y i ) (13) (14) (15) Here, ŷ i is y i β 0 β 1 x 1i β p x pi and β is [β 0 β 1 β p ]. As this objective function is based on median of absolute proportional error, it enables obtaining parameter estimations without taking outliers into consideration. As it is difficult to create normal equations in the optimization of given in (15), the optimization can be achieved with PSO. In prediction p + 1 parameter in PSO, the state of particle is expressed as in Fig. 1.

7 1274 Cagcag et al. gbest values obtained by employing algorithm 1 for the objective function in (15) are regression estimators. These estimators are express as follows. ˆβ PSO = [ ˆβ 0 ˆβ 1 ˆβ p ] (16) In the proposed method, the estimation of σ 2 is calculated as in Eq. (17). As this estimator relies on median, it will be less affected by the outlier in comparison with the OLS estimator given in (4) formula. ˆσ 2 PSO = Median( (Y Ŷ ) 2) (17) 5. Simulation Study for the Performance Evaluation of the Proposed Method A simulation study was designed to compare the proposed method with those found in the literature. In the simulation study, the performance of the methods was investigated for different values of the observation number and for σ 2. While evaluating the performance of the method, proximity to the true parameter value for β.dσ 2 parameter is calculated by the following criteria. MSE (β) = MSE (σ 2 ) = 1 tkr 1 (p + 1) tkr tkr j=1 p tkr (β ij ˆβ ij ) 2 (18) i=0 j=1 ( σ 2 j ˆσ j 2 ) 2 (19) The parameter of the method with small MSE value makes more accurate estimates. In Eqs (18) and (19), β.dσ 2 are real parameter values, ˆβ and ˆσ 2 are the parameter estimations obtained from simulation data. tkr represents the number of simulation data. To evaluate the performance of the method, below MSE values and the percentages of the method giving minimum error were used. MSE j (β) = 1 (p + 1) p (β i ˆβ i ) 2 j = 1, 2,...,tkr (20) i=0 MSE j (σ 2 ) = (σ 2 ˆσ 2 ) 2 j = 1, 2,...,tkr (21) For simulation, bivariate multiple regression model was designed. This model can be expressed by expressions (22) and (23). y i = β 0 + β 1 x 1i + β 2 x 2i + e i i = 1, 2,...,n (22) y i = 1 + 1x 1i + 1x 2i + e i i = 1, 2,...,n (23) Independent variables in the model were derived from a normal distribution with a mean of 100 and a variance of 100. Additionally, four different error data were created from normal distribution with a mean of 0 and a variance of 1, 9, 25, and 100, respectively. Moreover, four different data having these features with a sample size of 30, 50, 100, and 500 were created. Besides, data sets are transformed into data set with outliers by creating one, two, and three outliers with a value of 500. Simulation was performed with 10,000 iterations for each situation. At the end of the simulation, obtained results for the data sets

8 New Robust Regression Method Based on Particle Swarm Optimization 1275 Table 1 n = 30 and y 15 = 500 σ ˆβ 0 ˆβ 1 ˆβ 2 ˆσ 2 MSE (β) MSE (σ 2 ) SR (β) SR (σ 2 ) PROPOSED % % OLS % 0.00% LSR % 0.00% PROPOSED % % OLS % 0.00% LSR % 0.00% PROPOSED % % OLS % 0.00% LSR % 0.00% PROPOSED % % OLS % 0.00% LSR % 0.00% which are n = 30 and n = 50, and the success rates (SRs) of the method within the 10,000 iterations are shown in Tables 1 6. For the other situations, the SR of the methods is given in Tables 7and 8. A total of 48 different situations were analyzed in the simulation study. When the tables that summarize the results of the analyzed situations were considered, little estimation error was showed in proposed method for both model coefficients (β 0,β 1,β 2,,β p ) and variance of error, another parameter of the model, (σ 2 ) (see columns MSE(β) and MSE ( σ 2) of Tables 1 6). It can be seen that most of the 10,000 iterations of model coefficients (β 0,β 1,β 2,,β p ) in each situations are estimated with less error. Moreover, different simulation study was performed. We used chi-squared and F distributions that are known as skewed distributions. The aim of these simulations gives more evidence to test the efficiency of the proposed method. In the simulation, model can be expressed by (22) and (23) expressions. y i = β 0 + β 1 x 1i + β 2 x 2i + e i i = 1, 2,...,n (24) y i = 1 + 1x 1i + 1x 2i + e i i = 1, 2,...,n (25) Independent variables in the model were derived from a chi-squared distribution with two degrees of freedom and F distribution with 5 and five degrees of freedom. Additionally, four different error data were created from normal distribution with a mean of 0 and a Table 2 n = 50 and y 25 = 500 σ ˆβ 0 ˆβ 1 ˆβ 2 ˆσ 2 MSE (β) MSE (σ 2 ) SR (β) SR (σ 2 ) PROPOSED % % OLS % 0.00% LSR % 0.00% PROPOSED % % OLS % 0.00% LSR % 0.00% PROPOSED % % OLS % 0.00% LSR % 0.00% PROPOSED % % OLS % 0.00% LSR % 0.00%

9 1276 Cagcag et al. Table 3 n = 30 and y 5, y 25 = 500 σ ˆβ 0 ˆβ 1 ˆβ 2 ˆσ 2 MSE (β) MSE (σ 2 ) SR (β) SR (σ 2 ) PROPOSED % % OLS % 0.00% LSR % 0.00% PROPOSED % % OLS % 0.00% LSR % 0.00% PROPOSED % % OLS % 0.00% LSR % 0.00% PROPOSED % % OLS % 0.00% LSR % 0.00% Table 4 n = 50 and y 10,y 40 = 500 σ ˆβ 0 ˆβ 1 ˆβ 2 ˆσ 2 MSE (β) MSE (σ 2 ) SR (β) SR (σ 2 ) PROPOSED % % OLS % 0.00% LSR % 0.00% PROPOSED % % OLS % 0.00% LSR % 0.00% PROPOSED % % OLS % 0.00% LSR % 0.00% PROPOSED % % OLS % 0.00% LSR % 0.00% Table 5 n = 30 and y 5,y 15,y 25 = 500 σ ˆβ 0 ˆβ 1 ˆβ 2 ˆσ 2 MSE (β) MSE (σ 2 ) SR (β) SR (σ 2 ) PROPOSED % % OLS % 0.00% LSR % 0.00% PROPOSED % % OLS % 0.00% LSR % 0.00% PROPOSED % % OLS % 0.00% LSR % 0.00% PROPOSED % % OLS % 0.00% LSR % 0.00%

10 New Robust Regression Method Based on Particle Swarm Optimization 1277 Table 6 n = 50 and y 10,y 25,y 40 = 500 σ ˆβ 0 ˆβ 1 ˆβ 2 ˆσ 2 MSE (β) MSE (σ 2 ) SR (β) SR (σ 2 ) PROPOSED % % OLS % 0.00% LSR % 0.00% PROPOSED % % OLS % 0.00% LSR % 0.00% PROPOSED % % OLS % 0.00% LSR % 0.00% PROPOSED % % OLS % 0.00% LSR % 0.00% Table 7 Proportion of achievement in case of n = 100 y 50 = 500 y 25,y 75 = 500 y 25,y 50,y 75 = 500 σ SR (β) SR (σ 2 ) SR (β) SR (σ 2 ) SR (β) SR (σ 2 ) PROPOSED % % 90.90% % 93.10% % OLS 1.30% 0.00% 0.70% 0.00% 0.60% 0.00% LSR 12.10% 0.00% 8.40% 0.00% 6.30% 0.00% PROPOSED % % 91.30% % 92.30% % OLS 1.80% 0.00% 0.60% 0.00% 0.60% 0.00% LSR 9.40% 0.00% 8.10% 0.00% 7.10% 0.00% PROPOSED % % 93.60% % 94.40% % OLS 2.00% 0.00% 0.70% 0.00% 0.60% 0.00% LSR 6.90% 0.00% 5.70% 0.00% 5.00% 0.00% PROPOSED % % 93.80% % 95.70% % OLS 1.60% 0.00% 1.60% 0.00% 0.60% 0.00% LSR 4.60% 0.00% 4.60% 0.00% 3.70% 0.00% Table 8 Proportion of achievement in case of n = 500 y 250 = 500 y 100,y 400 = 500 y 100,y 250,y 400 = 500 σ SR (β) SR (σ 2 ) SR (β) SR (σ 2 ) SR (β) SR (σ 2 ) PROPOSED % % 61.00% % 68.40% % OLS 3.30% 0.00% 3.20% 0.00% 3.70% 0.00% LSR 47.00% 0.00% 35.80% 0.00% 27.90% 0.00% PROPOSED % % 68.40% % 72.90% % OLS 7.10% 0.00% 5.00% 0.00% 3.40% 0.00% LSR 29.70% 0.00% 26.60% 0.00% 23.70% 0.00% PROPOSED % % 77.30% % 75.90% % OLS 8.10% 0.00% 5.30% 0.00% 4.50% 0.00% LSR 19.50% 0.00% 17.40% 0.00% 19.60% 0.00% PROPOSED % % 83.60% % 86.70% % OLS 7.30% 0.00% 6.50% 0.00% 4.50% 0.00% LSR 10.30% 0.00% 9.90% 0.00% 8.80% 0.00%

11 1278 Cagcag et al. Table 9 Proportion of achievement in case of independent variables with chi-squared distribution One outlier Two outliers Three outliers σ SR (β) SR (σ 2 ) SR (β) SR (σ 2 ) SR (β) SR (σ 2 ) PROPOSED % % % % % % OLS 15.10% 0.00% 0.00% 0.00% 0.00% 0.00% LSR 0.10% 0.00% 0.00% 0.00% 0.00% 0.00% PROPOSED % % 99.70% % 99.20% % OLS 34.00% 0.00% 0.00% 0.00% 0.00% 0.00% LSR 0.10% 0.00% 0.30% 0.00% 0.80% 0.00% PROPOSED % % 94.40% % 93.30% % OLS 40.70% 0.00% 0.00% 0.00% 0.00% 0.00% LSR 0.40% 0.00% 5.60% 0.00% 6.70% 0.00% PROPOSED % % 71.40% % 71.30% % OLS 41.50% 0.00% 0.00% 0.00% 0.00% 0.00% LSR 8.70% 0.00% 28.60% 0.00% 28.70% 0.00% Table 10 Proportion of achievement in case of independent variables with F distribution One outlier Two outliers Three outliers σ SR (β) SR (σ 2 ) SR (β) SR (σ 2 ) SR (β) SR (σ 2 ) PROPOSED % % % % % % OLS 2.70% 0.00% 0.00% 0.00% 0.00% 0.00% LSR 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% PROPOSED % % 98.70% % 99.30% % OLS 22.60% 0.00% 0.00% 0.00% 0.00% 0.00% LSR 0.50% 0.00% 1.30% 0.00% 0.70% 0.00% PROPOSED % % 92.10% % 90.40% % OLS 35.80% 0.00% 0.00% 0.00% 0.00% 0.00% LSR 2.80% 0.00% 7.90% 0.00% 9.60% 0.00% PROPOSED % % 69.40% % 71.10% % OLS 43.80% 0.00% 0.00% 0.00% 0.00% 0.00% LSR 11.10% 0.00% 30.60% 0.00% 28.90% 0.00% Table 11 Sample data containing outlier in its dependent variable n i x 1i x 2i x 3i y i n i x 1i x 2i x 3i y i

12 New Robust Regression Method Based on Particle Swarm Optimization 1279 Table 12 Results of various regression methods for sample data ˆβ 0 ˆβ 1 ˆβ 2 ˆβ 3 MDAPE OLS MESTIMATOR Huber Hampel Tukey Andrews Sanli LSR Proposed variance of 1, 9, 25, and 100, respectively. Moreover, data having these features with a sample size of 500 were created. Besides, data sets are transformed into data set with outliers by creating one, two, and three outliers. Outlier values were obtained by multiplying the maximum observation value of the original data by 10. Then, 250th, 100th and 400th, 100th, 250th and 400th observation of the original data, which is the maximum observation value, were changed by these outliers. Simulation was performed with 10,000 iterations for each situation. At the end of the simulation, obtained results for the data sets are shown in Tables 9 and 10. To strengthen the superior performance of the proposed method, in addition to the simulation study, another analysis was done on a single data that was given in Table 11, and results were compared with the other robust methods. where X 1 (μ = 20,σ = 3), X 2 (μ = 50,σ = 12), X 3 (μ = 32,σ = 13). Dependent variables are created as below but 15th observation in dependent variable was transformed into outlier to be as y y i = 1 + 1x 1i + 1x 2i + 1x 3i + e i i = 1, 2,...,n (26) Additionally, the error data was created from normal distribution with a mean of 0 and a variance of 9. The results of the proposed method as well as the results of some robust methods were summarized in Table 12. In this application, MDAPE criterion was used in the comparison of methods. As MDAPE criterion is based on median, its use will be more reliable in comparing methods which are used for the analysis of data containing outlier. ( ) y i ŷ i MDAPE = median, i = 1, 2,,n (27) y i When Table 12 was analyzed, it was evident that the proposed method was superior to other methods in terms of MDAPE criterion. 6. Conclusions and Discussion OLS is one of the most frequently used methods in regression analysis which are most frequently referred in estimation problems. But, being sensitive to the outliers in the data

13 1280 Cagcag et al. led to the researchers to put forward alternative methods known as robust methods. M- estimators, one of the robust methods, focus on minimizing a function of residuals rather than the squares of residuals. Additionally, those methods which are based on simple fuzzy regression analysis and minimizing the proportional error rather than absolute error are also available in literature. In this study, a new method minimizing the median of proportional error with PSO in analyzing data containing outlier was proposed. The performance of the proposed method was applied to a simulation study, and a derived single data as well and obtained results were evaluated. When the results obtained from the application were analyzed, it was seen that proportional error was minimized with PSO and the method displayed superior performance when compared with the some other robust methods. The study results also revealed that estimators in the proposed method were more effective and unbiased in comparison with the OLS method. Although, the proposed method provides more biased results in comparison with the LSR method, it is seen that the proposed method is a better estimator. Therefore, it can be concluded that this method provides better estimation results in regard to both LSR and OLS methods. References Akbilgic, O., Akinci, E. D. (2009). A novel regression approach: Least squares ratio. Commun. Stat. Theory Methods 38: Candan, M. (1995). Robust estimator in linear regression. Ph.D. Thesis, Hacettepe University Ankara. Dave, R.N., Krishnapuram, R. (1997). Robust clustering methods: A unified view. IEEE Trans. Fuzzy Syst. 5(2): Hampel, F. R., Ronchetti, E. M., Rousseeuw, P. J., Shatel, W. A. (1986). Robust Statistics (pp , 502 p). New York: John Wiley & Sons. Hogg, R. V. (1979). Statistical robustness: One view of its use in applications today. Am. Stat. 33(3): Huber, P. J. (1981). Robust Statistics (pp. 1 20, , 308 p). New York: John Wiley & Sons. Huynh, H. (1982). A comparison of four approaches to robust regression. Psychol. Bull. 92(2): Kennedy, J., Eberhart, R. (1995). Particle swarm optimization. In Proceedings of IEEE International Conference on Neural Networks, pp , Piscataway, NJ: IEEE Press. Ma, Y., Jiang, C., Hou, Z., Wang, C. (2006). The formulation of the optimal strategies for the electricity producers based on the particle swarm optimization algorithm. IEEE Trans. Power Syst. 21(4): Rousseuw, P. J., Leroy, A. M. (1997). Robust Regression and Outlier Detection (pp , 329 p.). New York: John Wiley & Sons. Sanli, K. (2005). Fuzzy robust regression analysis, Ph.D. Thesis. Ankara University, Ankara. Shi, Y., Eberhart, R. C. (1999). Empirical study of particle swarm optimization. Proc. IEEE Int. Congr. Evol. Comput. 3:

Published online: 17 Jun 2010.

Published online: 17 Jun 2010. This article was downloaded by: [Sam Houston State University] On: 07 August 2014, At: 15:09 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered

More information

Online publication date: 19 May 2010 PLEASE SCROLL DOWN FOR ARTICLE

Online publication date: 19 May 2010 PLEASE SCROLL DOWN FOR ARTICLE This article was downloaded by: [Patterson, David A.] On: 19 May 2010 Access details: Access Details: [subscription number 922426156] Publisher Routledge Informa Ltd Registered in England and Wales Registered

More information

Optimization of PID parameters with an improved simplex PSO

Optimization of PID parameters with an improved simplex PSO Li et al. Journal of Inequalities and Applications (2015) 2015:325 DOI 10.1186/s13660-015-0785-2 R E S E A R C H Open Access Optimization of PID parameters with an improved simplex PSO Ji-min Li 1, Yeong-Cheng

More information

Daring Greatly: How the Courage to Be Vulnerable Transforms the Way We Live, Love, Parent, and Lead. Click for updates

Daring Greatly: How the Courage to Be Vulnerable Transforms the Way We Live, Love, Parent, and Lead. Click for updates This article was downloaded by: [184.100.72.114] On: 19 January 2015, At: 17:22 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,

More information

Nilpotent Lie and Leibniz Algebras

Nilpotent Lie and Leibniz Algebras This article was downloaded by: [North Carolina State University] On: 03 March 2014, At: 08:05 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered

More information

Overview of Violations of the Basic Assumptions in the Classical Normal Linear Regression Model

Overview of Violations of the Basic Assumptions in the Classical Normal Linear Regression Model Overview of Violations of the Basic Assumptions in the Classical Normal Linear Regression Model 1 September 004 A. Introduction and assumptions The classical normal linear regression model can be written

More information

Beijing, China b CMOE Key Laboratory of Petroleum Engineering in China University

Beijing, China b CMOE Key Laboratory of Petroleum Engineering in China University This article was downloaded by: [Zhejiang University On: 21 September 2014, At: 03:04 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office:

More information

The Prevalence and Prevention of Crosstalk: A Multi-Institutional Study

The Prevalence and Prevention of Crosstalk: A Multi-Institutional Study This article was downloaded by: [65.186.78.206] On: 10 April 2014, At: 17:16 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,

More information

PLEASE SCROLL DOWN FOR ARTICLE. Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf

PLEASE SCROLL DOWN FOR ARTICLE. Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf This article was downloaded by: On: 6 January 2010 Access details: Access Details: Free Access Publisher Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered

More information

Least Squares Estimation

Least Squares Estimation Least Squares Estimation SARA A VAN DE GEER Volume 2, pp 1041 1045 in Encyclopedia of Statistics in Behavioral Science ISBN-13: 978-0-470-86080-9 ISBN-10: 0-470-86080-4 Editors Brian S Everitt & David

More information

A New Quantitative Behavioral Model for Financial Prediction

A New Quantitative Behavioral Model for Financial Prediction 2011 3rd International Conference on Information and Financial Engineering IPEDR vol.12 (2011) (2011) IACSIT Press, Singapore A New Quantitative Behavioral Model for Financial Prediction Thimmaraya Ramesh

More information

Rens van de Schoot a b, Peter Lugtig a & Joop Hox a a Department of Methods and Statistics, Utrecht

Rens van de Schoot a b, Peter Lugtig a & Joop Hox a a Department of Methods and Statistics, Utrecht This article was downloaded by: [University Library Utrecht] On: 15 May 2012, At: 01:20 Publisher: Psychology Press Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office:

More information

Optimal PID Controller Design for AVR System

Optimal PID Controller Design for AVR System Tamkang Journal of Science and Engineering, Vol. 2, No. 3, pp. 259 270 (2009) 259 Optimal PID Controller Design for AVR System Ching-Chang Wong*, Shih-An Li and Hou-Yi Wang Department of Electrical Engineering,

More information

CLOUD DATABASE ROUTE SCHEDULING USING COMBANATION OF PARTICLE SWARM OPTIMIZATION AND GENETIC ALGORITHM

CLOUD DATABASE ROUTE SCHEDULING USING COMBANATION OF PARTICLE SWARM OPTIMIZATION AND GENETIC ALGORITHM CLOUD DATABASE ROUTE SCHEDULING USING COMBANATION OF PARTICLE SWARM OPTIMIZATION AND GENETIC ALGORITHM *Shabnam Ghasemi 1 and Mohammad Kalantari 2 1 Deparment of Computer Engineering, Islamic Azad University,

More information

A Novel Binary Particle Swarm Optimization

A Novel Binary Particle Swarm Optimization Proceedings of the 5th Mediterranean Conference on T33- A Novel Binary Particle Swarm Optimization Motaba Ahmadieh Khanesar, Member, IEEE, Mohammad Teshnehlab and Mahdi Aliyari Shoorehdeli K. N. Toosi

More information

California Published online: 09 Jun 2014.

California Published online: 09 Jun 2014. This article was downloaded by: [Mr Neil Ribner] On: 10 June 2014, At: 20:58 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,

More information

A Comparative Study of the Pickup Method and its Variations Using a Simulated Hotel Reservation Data

A Comparative Study of the Pickup Method and its Variations Using a Simulated Hotel Reservation Data A Comparative Study of the Pickup Method and its Variations Using a Simulated Hotel Reservation Data Athanasius Zakhary, Neamat El Gayar Faculty of Computers and Information Cairo University, Giza, Egypt

More information

Multiple Linear Regression in Data Mining

Multiple Linear Regression in Data Mining Multiple Linear Regression in Data Mining Contents 2.1. A Review of Multiple Linear Regression 2.2. Illustration of the Regression Process 2.3. Subset Selection in Linear Regression 1 2 Chap. 2 Multiple

More information

STATISTICS AND DATA ANALYSIS IN GEOLOGY, 3rd ed. Clarificationof zonationprocedure described onpp. 238-239

STATISTICS AND DATA ANALYSIS IN GEOLOGY, 3rd ed. Clarificationof zonationprocedure described onpp. 238-239 STATISTICS AND DATA ANALYSIS IN GEOLOGY, 3rd ed. by John C. Davis Clarificationof zonationprocedure described onpp. 38-39 Because the notation used in this section (Eqs. 4.8 through 4.84) is inconsistent

More information

How To Understand The History Of Part Time Business Studies

How To Understand The History Of Part Time Business Studies This article was downloaded by: [148.251.235.206] On: 27 August 2015, At: 06:33 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: 5 Howick Place,

More information

A RANDOMIZED LOAD BALANCING ALGORITHM IN GRID USING MAX MIN PSO ALGORITHM

A RANDOMIZED LOAD BALANCING ALGORITHM IN GRID USING MAX MIN PSO ALGORITHM International Journal of Research in Computer Science eissn 2249-8265 Volume 2 Issue 3 (212) pp. 17-23 White Globe Publications A RANDOMIZED LOAD BALANCING ALGORITHM IN GRID USING MAX MIN ALGORITHM C.Kalpana

More information

Business Security Architecture: Weaving Information Security into Your Organization's Enterprise Architecture through SABSA

Business Security Architecture: Weaving Information Security into Your Organization's Enterprise Architecture through SABSA This article was downloaded by: [188.204.15.66] On: 20 February 2012, At: 01:40 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer

More information

BMOA: Binary Magnetic Optimization Algorithm

BMOA: Binary Magnetic Optimization Algorithm International Journal of Machine Learning and Computing Vol. 2 No. 3 June 22 BMOA: Binary Magnetic Optimization Algorithm SeyedAli Mirjalili and Siti Zaiton Mohd Hashim Abstract Recently the behavior of

More information

Simple linear regression

Simple linear regression Simple linear regression Introduction Simple linear regression is a statistical method for obtaining a formula to predict values of one variable from another where there is a causal relationship between

More information

Optimal Tuning of PID Controller Using Meta Heuristic Approach

Optimal Tuning of PID Controller Using Meta Heuristic Approach International Journal of Electronic and Electrical Engineering. ISSN 0974-2174, Volume 7, Number 2 (2014), pp. 171-176 International Research Publication House http://www.irphouse.com Optimal Tuning of

More information

NASPE Sets the Standard

NASPE Sets the Standard This article was downloaded by: [Bowling Green SU] On: 25 March 2015, At: 09:45 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,

More information

Applied Mathematical Sciences, Vol. 7, 2013, no. 112, 5591-5597 HIKARI Ltd, www.m-hikari.com http://dx.doi.org/10.12988/ams.2013.

Applied Mathematical Sciences, Vol. 7, 2013, no. 112, 5591-5597 HIKARI Ltd, www.m-hikari.com http://dx.doi.org/10.12988/ams.2013. Applied Mathematical Sciences, Vol. 7, 2013, no. 112, 5591-5597 HIKARI Ltd, www.m-hikari.com http://dx.doi.org/10.12988/ams.2013.38457 Accuracy Rate of Predictive Models in Credit Screening Anirut Suebsing

More information

Using Learning from Work for Progression to Higher Education: a degree of experience

Using Learning from Work for Progression to Higher Education: a degree of experience This article was downloaded by: [148.251.235.206] On: 27 August 2015, At: 21:16 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: 5 Howick Place,

More information

The Basics of FEA Procedure

The Basics of FEA Procedure CHAPTER 2 The Basics of FEA Procedure 2.1 Introduction This chapter discusses the spring element, especially for the purpose of introducing various concepts involved in use of the FEA technique. A spring

More information

Imputing Missing Data using SAS

Imputing Missing Data using SAS ABSTRACT Paper 3295-2015 Imputing Missing Data using SAS Christopher Yim, California Polytechnic State University, San Luis Obispo Missing data is an unfortunate reality of statistics. However, there are

More information

Fairfield Public Schools

Fairfield Public Schools Mathematics Fairfield Public Schools AP Statistics AP Statistics BOE Approved 04/08/2014 1 AP STATISTICS Critical Areas of Focus AP Statistics is a rigorous course that offers advanced students an opportunity

More information

THREE DIMENSIONAL REPRESENTATION OF AMINO ACID CHARAC- TERISTICS

THREE DIMENSIONAL REPRESENTATION OF AMINO ACID CHARAC- TERISTICS THREE DIMENSIONAL REPRESENTATION OF AMINO ACID CHARAC- TERISTICS O.U. Sezerman 1, R. Islamaj 2, E. Alpaydin 2 1 Laborotory of Computational Biology, Sabancı University, Istanbul, Turkey. 2 Computer Engineering

More information

Univariate Regression

Univariate Regression Univariate Regression Correlation and Regression The regression line summarizes the linear relationship between 2 variables Correlation coefficient, r, measures strength of relationship: the closer r is

More information

Dynamic Generation of Test Cases with Metaheuristics

Dynamic Generation of Test Cases with Metaheuristics Dynamic Generation of Test Cases with Metaheuristics Laura Lanzarini, Juan Pablo La Battaglia III-LIDI (Institute of Research in Computer Science LIDI) Faculty of Computer Sciences. National University

More information

Wireless Sensor Networks Coverage Optimization based on Improved AFSA Algorithm

Wireless Sensor Networks Coverage Optimization based on Improved AFSA Algorithm , pp. 99-108 http://dx.doi.org/10.1457/ijfgcn.015.8.1.11 Wireless Sensor Networks Coverage Optimization based on Improved AFSA Algorithm Wang DaWei and Wang Changliang Zhejiang Industry Polytechnic College

More information

STATISTICA Formula Guide: Logistic Regression. Table of Contents

STATISTICA Formula Guide: Logistic Regression. Table of Contents : Table of Contents... 1 Overview of Model... 1 Dispersion... 2 Parameterization... 3 Sigma-Restricted Model... 3 Overparameterized Model... 4 Reference Coding... 4 Model Summary (Summary Tab)... 5 Summary

More information

Web-based Supplementary Materials for Bayesian Effect Estimation. Accounting for Adjustment Uncertainty by Chi Wang, Giovanni

Web-based Supplementary Materials for Bayesian Effect Estimation. Accounting for Adjustment Uncertainty by Chi Wang, Giovanni 1 Web-based Supplementary Materials for Bayesian Effect Estimation Accounting for Adjustment Uncertainty by Chi Wang, Giovanni Parmigiani, and Francesca Dominici In Web Appendix A, we provide detailed

More information

INTERPRETING THE ONE-WAY ANALYSIS OF VARIANCE (ANOVA)

INTERPRETING THE ONE-WAY ANALYSIS OF VARIANCE (ANOVA) INTERPRETING THE ONE-WAY ANALYSIS OF VARIANCE (ANOVA) As with other parametric statistics, we begin the one-way ANOVA with a test of the underlying assumptions. Our first assumption is the assumption of

More information

Predictability of Non-Linear Trading Rules in the US Stock Market Chong & Lam 2010

Predictability of Non-Linear Trading Rules in the US Stock Market Chong & Lam 2010 Department of Mathematics QF505 Topics in quantitative finance Group Project Report Predictability of on-linear Trading Rules in the US Stock Market Chong & Lam 010 ame: Liu Min Qi Yichen Zhang Fengtian

More information

International Journal of Software and Web Sciences (IJSWS) www.iasir.net

International Journal of Software and Web Sciences (IJSWS) www.iasir.net International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research) ISSN (Print): 2279-0063 ISSN (Online): 2279-0071 International

More information

University of Ljubljana Doctoral Programme in Statistics Methodology of Statistical Research Written examination February 14 th, 2014.

University of Ljubljana Doctoral Programme in Statistics Methodology of Statistical Research Written examination February 14 th, 2014. University of Ljubljana Doctoral Programme in Statistics ethodology of Statistical Research Written examination February 14 th, 2014 Name and surname: ID number: Instructions Read carefully the wording

More information

Keywords: Image complexity, PSNR, Levenberg-Marquardt, Multi-layer neural network.

Keywords: Image complexity, PSNR, Levenberg-Marquardt, Multi-layer neural network. Global Journal of Computer Science and Technology Volume 11 Issue 3 Version 1.0 Type: Double Blind Peer Reviewed International Research Journal Publisher: Global Journals Inc. (USA) Online ISSN: 0975-4172

More information

A COMPARISON OF REGRESSION MODELS FOR FORECASTING A CUMULATIVE VARIABLE

A COMPARISON OF REGRESSION MODELS FOR FORECASTING A CUMULATIVE VARIABLE A COMPARISON OF REGRESSION MODELS FOR FORECASTING A CUMULATIVE VARIABLE Joanne S. Utley, School of Business and Economics, North Carolina A&T State University, Greensboro, NC 27411, (336)-334-7656 (ext.

More information

Statistical Models in R

Statistical Models in R Statistical Models in R Some Examples Steven Buechler Department of Mathematics 276B Hurley Hall; 1-6233 Fall, 2007 Outline Statistical Models Structure of models in R Model Assessment (Part IA) Anova

More information

Quartile-Based Defuzzify Scheme Integrated with PSO for Revenue Change Forecasting. Abstract

Quartile-Based Defuzzify Scheme Integrated with PSO for Revenue Change Forecasting. Abstract Quartile-Based Defuzzify Scheme Integrated with PSO for Revenue Change Forecasting Yao-Lin Huang 1 I-Hong Kuo 2 Shu-Mei Pai 3 1 Department of Computer and Communication Engineering, De Lin Institute of

More information

PARTIAL LEAST SQUARES IS TO LISREL AS PRINCIPAL COMPONENTS ANALYSIS IS TO COMMON FACTOR ANALYSIS. Wynne W. Chin University of Calgary, CANADA

PARTIAL LEAST SQUARES IS TO LISREL AS PRINCIPAL COMPONENTS ANALYSIS IS TO COMMON FACTOR ANALYSIS. Wynne W. Chin University of Calgary, CANADA PARTIAL LEAST SQUARES IS TO LISREL AS PRINCIPAL COMPONENTS ANALYSIS IS TO COMMON FACTOR ANALYSIS. Wynne W. Chin University of Calgary, CANADA ABSTRACT The decision of whether to use PLS instead of a covariance

More information

Time Series and Forecasting

Time Series and Forecasting Chapter 22 Page 1 Time Series and Forecasting A time series is a sequence of observations of a random variable. Hence, it is a stochastic process. Examples include the monthly demand for a product, the

More information

Biostatistics: DESCRIPTIVE STATISTICS: 2, VARIABILITY

Biostatistics: DESCRIPTIVE STATISTICS: 2, VARIABILITY Biostatistics: DESCRIPTIVE STATISTICS: 2, VARIABILITY 1. Introduction Besides arriving at an appropriate expression of an average or consensus value for observations of a population, it is important to

More information

5. Linear Regression

5. Linear Regression 5. Linear Regression Outline.................................................................... 2 Simple linear regression 3 Linear model............................................................. 4

More information

Factor Analysis. Chapter 420. Introduction

Factor Analysis. Chapter 420. Introduction Chapter 420 Introduction (FA) is an exploratory technique applied to a set of observed variables that seeks to find underlying factors (subsets of variables) from which the observed variables were generated.

More information

Demand Forecasting Optimization in Supply Chain

Demand Forecasting Optimization in Supply Chain 2011 International Conference on Information Management and Engineering (ICIME 2011) IPCSIT vol. 52 (2012) (2012) IACSIT Press, Singapore DOI: 10.7763/IPCSIT.2012.V52.12 Demand Forecasting Optimization

More information

A Divided Regression Analysis for Big Data

A Divided Regression Analysis for Big Data Vol., No. (0), pp. - http://dx.doi.org/0./ijseia.0...0 A Divided Regression Analysis for Big Data Sunghae Jun, Seung-Joo Lee and Jea-Bok Ryu Department of Statistics, Cheongju University, 0-, Korea shjun@cju.ac.kr,

More information

XOR-based artificial bee colony algorithm for binary optimization

XOR-based artificial bee colony algorithm for binary optimization Turkish Journal of Electrical Engineering & Computer Sciences http:// journals. tubitak. gov. tr/ elektrik/ Research Article Turk J Elec Eng & Comp Sci (2013) 21: 2307 2328 c TÜBİTAK doi:10.3906/elk-1203-104

More information

Online publication date: 15 March 2010

Online publication date: 15 March 2010 This article was downloaded by: [Swets Content Distribution] On: 17 September 2010 Access details: Access Details: [subscription number 925215345] Publisher Routledge Informa Ltd Registered in England

More information

CS 147: Computer Systems Performance Analysis

CS 147: Computer Systems Performance Analysis CS 147: Computer Systems Performance Analysis One-Factor Experiments CS 147: Computer Systems Performance Analysis One-Factor Experiments 1 / 42 Overview Introduction Overview Overview Introduction Finding

More information

Simple Linear Regression Inference

Simple Linear Regression Inference Simple Linear Regression Inference 1 Inference requirements The Normality assumption of the stochastic term e is needed for inference even if it is not a OLS requirement. Therefore we have: Interpretation

More information

430 Statistics and Financial Mathematics for Business

430 Statistics and Financial Mathematics for Business Prescription: 430 Statistics and Financial Mathematics for Business Elective prescription Level 4 Credit 20 Version 2 Aim Students will be able to summarise, analyse, interpret and present data, make predictions

More information

Multivariate Analysis of Variance (MANOVA)

Multivariate Analysis of Variance (MANOVA) Multivariate Analysis of Variance (MANOVA) Aaron French, Marcelo Macedo, John Poulsen, Tyler Waterson and Angela Yu Keywords: MANCOVA, special cases, assumptions, further reading, computations Introduction

More information

Comparison of sales forecasting models for an innovative agro-industrial product: Bass model versus logistic function

Comparison of sales forecasting models for an innovative agro-industrial product: Bass model versus logistic function The Empirical Econometrics and Quantitative Economics Letters ISSN 2286 7147 EEQEL all rights reserved Volume 1, Number 4 (December 2012), pp. 89 106. Comparison of sales forecasting models for an innovative

More information

Optimized Fuzzy Control by Particle Swarm Optimization Technique for Control of CSTR

Optimized Fuzzy Control by Particle Swarm Optimization Technique for Control of CSTR International Journal of Computer, Electrical, Automation, Control and Information Engineering Vol:5, No:, 20 Optimized Fuzzy Control by Particle Swarm Optimization Technique for Control of CSTR Saeed

More information

Chapter 6. The stacking ensemble approach

Chapter 6. The stacking ensemble approach 82 This chapter proposes the stacking ensemble approach for combining different data mining classifiers to get better performance. Other combination techniques like voting, bagging etc are also described

More information

Robust procedures for Canadian Test Day Model final report for the Holstein breed

Robust procedures for Canadian Test Day Model final report for the Holstein breed Robust procedures for Canadian Test Day Model final report for the Holstein breed J. Jamrozik, J. Fatehi and L.R. Schaeffer Centre for Genetic Improvement of Livestock, University of Guelph Introduction

More information

Introduction to Matrix Algebra

Introduction to Matrix Algebra Psychology 7291: Multivariate Statistics (Carey) 8/27/98 Matrix Algebra - 1 Introduction to Matrix Algebra Definitions: A matrix is a collection of numbers ordered by rows and columns. It is customary

More information

Standardization and Its Effects on K-Means Clustering Algorithm

Standardization and Its Effects on K-Means Clustering Algorithm Research Journal of Applied Sciences, Engineering and Technology 6(7): 399-3303, 03 ISSN: 040-7459; e-issn: 040-7467 Maxwell Scientific Organization, 03 Submitted: January 3, 03 Accepted: February 5, 03

More information

PREDICTION FOR SHORT-TERM TRAFFIC FLOW BASED ON OPTIMIZED WAVELET NEURAL NETWORK MODEL

PREDICTION FOR SHORT-TERM TRAFFIC FLOW BASED ON OPTIMIZED WAVELET NEURAL NETWORK MODEL PREDICTION FOR SHORT-TERM TRAFFIC FLOW BASED ON OPTIMIZED WAVELET NEURAL NETWORK MODEL ABSTRACT Tao Li 1 and Liu Sheng 2 School of Management, Shanghai University of Engineering Science Shanghai, China

More information

Permutation Tests for Comparing Two Populations

Permutation Tests for Comparing Two Populations Permutation Tests for Comparing Two Populations Ferry Butar Butar, Ph.D. Jae-Wan Park Abstract Permutation tests for comparing two populations could be widely used in practice because of flexibility of

More information

By choosing to view this document, you agree to all provisions of the copyright laws protecting it.

By choosing to view this document, you agree to all provisions of the copyright laws protecting it. This material is posted here with permission of the IEEE Such permission of the IEEE does not in any way imply IEEE endorsement of any of Helsinki University of Technology's products or services Internal

More information

CALCULATIONS & STATISTICS

CALCULATIONS & STATISTICS CALCULATIONS & STATISTICS CALCULATION OF SCORES Conversion of 1-5 scale to 0-100 scores When you look at your report, you will notice that the scores are reported on a 0-100 scale, even though respondents

More information

APPLICATION OF MODIFIED (PSO) AND SIMULATED ANNEALING ALGORITHM (SAA) IN ECONOMIC LOAD DISPATCH PROBLEM OF THERMAL GENERATING UNIT

APPLICATION OF MODIFIED (PSO) AND SIMULATED ANNEALING ALGORITHM (SAA) IN ECONOMIC LOAD DISPATCH PROBLEM OF THERMAL GENERATING UNIT International Journal of Electrical Engineering & Technology (IJEET) Volume 7, Issue 2, March-April, 2016, pp.69 78, Article ID: IJEET_07_02_008 Available online at http:// http://www.iaeme.com/ijeet/issues.asp?jtype=ijeet&vtype=7&itype=2

More information

Content Sheet 7-1: Overview of Quality Control for Quantitative Tests

Content Sheet 7-1: Overview of Quality Control for Quantitative Tests Content Sheet 7-1: Overview of Quality Control for Quantitative Tests Role in quality management system Quality Control (QC) is a component of process control, and is a major element of the quality management

More information

Institute of Actuaries of India Subject CT3 Probability and Mathematical Statistics

Institute of Actuaries of India Subject CT3 Probability and Mathematical Statistics Institute of Actuaries of India Subject CT3 Probability and Mathematical Statistics For 2015 Examinations Aim The aim of the Probability and Mathematical Statistics subject is to provide a grounding in

More information

Forecasting the sales of an innovative agro-industrial product with limited information: A case of feta cheese from buffalo milk in Thailand

Forecasting the sales of an innovative agro-industrial product with limited information: A case of feta cheese from buffalo milk in Thailand Forecasting the sales of an innovative agro-industrial product with limited information: A case of feta cheese from buffalo milk in Thailand Orakanya Kanjanatarakul 1 and Komsan Suriya 2 1 Faculty of Economics,

More information

Chapter 13 Introduction to Nonlinear Regression( 非 線 性 迴 歸 )

Chapter 13 Introduction to Nonlinear Regression( 非 線 性 迴 歸 ) Chapter 13 Introduction to Nonlinear Regression( 非 線 性 迴 歸 ) and Neural Networks( 類 神 經 網 路 ) 許 湘 伶 Applied Linear Regression Models (Kutner, Nachtsheim, Neter, Li) hsuhl (NUK) LR Chap 10 1 / 35 13 Examples

More information

Mehtap Ergüven Abstract of Ph.D. Dissertation for the degree of PhD of Engineering in Informatics

Mehtap Ergüven Abstract of Ph.D. Dissertation for the degree of PhD of Engineering in Informatics INTERNATIONAL BLACK SEA UNIVERSITY COMPUTER TECHNOLOGIES AND ENGINEERING FACULTY ELABORATION OF AN ALGORITHM OF DETECTING TESTS DIMENSIONALITY Mehtap Ergüven Abstract of Ph.D. Dissertation for the degree

More information

Current Standard: Mathematical Concepts and Applications Shape, Space, and Measurement- Primary

Current Standard: Mathematical Concepts and Applications Shape, Space, and Measurement- Primary Shape, Space, and Measurement- Primary A student shall apply concepts of shape, space, and measurement to solve problems involving two- and three-dimensional shapes by demonstrating an understanding of:

More information

Sections 2.11 and 5.8

Sections 2.11 and 5.8 Sections 211 and 58 Timothy Hanson Department of Statistics, University of South Carolina Stat 704: Data Analysis I 1/25 Gesell data Let X be the age in in months a child speaks his/her first word and

More information

Patent Big Data Analysis by R Data Language for Technology Management

Patent Big Data Analysis by R Data Language for Technology Management , pp. 69-78 http://dx.doi.org/10.14257/ijseia.2016.10.1.08 Patent Big Data Analysis by R Data Language for Technology Management Sunghae Jun * Department of Statistics, Cheongju University, 360-764, Korea

More information

POLYNOMIAL AND MULTIPLE REGRESSION. Polynomial regression used to fit nonlinear (e.g. curvilinear) data into a least squares linear regression model.

POLYNOMIAL AND MULTIPLE REGRESSION. Polynomial regression used to fit nonlinear (e.g. curvilinear) data into a least squares linear regression model. Polynomial Regression POLYNOMIAL AND MULTIPLE REGRESSION Polynomial regression used to fit nonlinear (e.g. curvilinear) data into a least squares linear regression model. It is a form of linear regression

More information

Random variables, probability distributions, binomial random variable

Random variables, probability distributions, binomial random variable Week 4 lecture notes. WEEK 4 page 1 Random variables, probability distributions, binomial random variable Eample 1 : Consider the eperiment of flipping a fair coin three times. The number of tails that

More information

Applications of improved grey prediction model for power demand forecasting

Applications of improved grey prediction model for power demand forecasting Energy Conversion and Management 44 (2003) 2241 2249 www.elsevier.com/locate/enconman Applications of improved grey prediction model for power demand forecasting Che-Chiang Hsu a, *, Chia-Yon Chen b a

More information

Predict the Popularity of YouTube Videos Using Early View Data

Predict the Popularity of YouTube Videos Using Early View Data 000 001 002 003 004 005 006 007 008 009 010 011 012 013 014 015 016 017 018 019 020 021 022 023 024 025 026 027 028 029 030 031 032 033 034 035 036 037 038 039 040 041 042 043 044 045 046 047 048 049 050

More information

USING COMPUTING INTELLIGENCE TECHNIQUES TO ESTIMATE SOFTWARE EFFORT

USING COMPUTING INTELLIGENCE TECHNIQUES TO ESTIMATE SOFTWARE EFFORT USING COMPUTING INTELLIGENCE TECHNIQUES TO ESTIMATE SOFTWARE EFFORT Jin-Cherng Lin, Yueh-Ting Lin, Han-Yuan Tzeng and Yan-Chin Wang Dept. of Computer Science & Engineering Tatung University Taipei 10452,

More information

MATHEMATICAL METHODS OF STATISTICS

MATHEMATICAL METHODS OF STATISTICS MATHEMATICAL METHODS OF STATISTICS By HARALD CRAMER TROFESSOK IN THE UNIVERSITY OF STOCKHOLM Princeton PRINCETON UNIVERSITY PRESS 1946 TABLE OF CONTENTS. First Part. MATHEMATICAL INTRODUCTION. CHAPTERS

More information

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions This article was downloaded by: [Lanzhou Institute of Geology] On: 27 February 2013, At: 01:00 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered

More information

SPECIAL PERTURBATIONS UNCORRELATED TRACK PROCESSING

SPECIAL PERTURBATIONS UNCORRELATED TRACK PROCESSING AAS 07-228 SPECIAL PERTURBATIONS UNCORRELATED TRACK PROCESSING INTRODUCTION James G. Miller * Two historical uncorrelated track (UCT) processing approaches have been employed using general perturbations

More information

Improved Particle Swarm Optimization in Constrained Numerical Search Spaces

Improved Particle Swarm Optimization in Constrained Numerical Search Spaces Improved Particle Swarm Optimization in Constrained Numerical Search Spaces Efrén Mezura-Montes and Jorge Isacc Flores-Mendoza Abstract This chapter presents a study about the behavior of Particle Swarm

More information

Research on the Performance Optimization of Hadoop in Big Data Environment

Research on the Performance Optimization of Hadoop in Big Data Environment Vol.8, No.5 (015), pp.93-304 http://dx.doi.org/10.1457/idta.015.8.5.6 Research on the Performance Optimization of Hadoop in Big Data Environment Jia Min-Zheng Department of Information Engineering, Beiing

More information

A Hybrid Model of Particle Swarm Optimization (PSO) and Artificial Bee Colony (ABC) Algorithm for Test Case Optimization

A Hybrid Model of Particle Swarm Optimization (PSO) and Artificial Bee Colony (ABC) Algorithm for Test Case Optimization A Hybrid Model of Particle Swarm Optimization (PSO) and Artificial Bee Colony (ABC) Algorithm for Test Case Optimization Abraham Kiran Joseph a, Dr. G. Radhamani b * a Research Scholar, Dr.G.R Damodaran

More information

Enhancing the SNR of the Fiber Optic Rotation Sensor using the LMS Algorithm

Enhancing the SNR of the Fiber Optic Rotation Sensor using the LMS Algorithm 1 Enhancing the SNR of the Fiber Optic Rotation Sensor using the LMS Algorithm Hani Mehrpouyan, Student Member, IEEE, Department of Electrical and Computer Engineering Queen s University, Kingston, Ontario,

More information

Logistic Regression. Jia Li. Department of Statistics The Pennsylvania State University. Logistic Regression

Logistic Regression. Jia Li. Department of Statistics The Pennsylvania State University. Logistic Regression Logistic Regression Department of Statistics The Pennsylvania State University Email: jiali@stat.psu.edu Logistic Regression Preserve linear classification boundaries. By the Bayes rule: Ĝ(x) = arg max

More information

Dynamic Task Scheduling with Load Balancing using Hybrid Particle Swarm Optimization

Dynamic Task Scheduling with Load Balancing using Hybrid Particle Swarm Optimization Int. J. Open Problems Compt. Math., Vol. 2, No. 3, September 2009 ISSN 1998-6262; Copyright ICSRS Publication, 2009 www.i-csrs.org Dynamic Task Scheduling with Load Balancing using Hybrid Particle Swarm

More information

Empirical Model-Building and Response Surfaces

Empirical Model-Building and Response Surfaces Empirical Model-Building and Response Surfaces GEORGE E. P. BOX NORMAN R. DRAPER Technische Universitat Darmstadt FACHBEREICH INFORMATIK BIBLIOTHEK Invortar-Nf.-. Sachgsbiete: Standort: New York John Wiley

More information

Follow links Class Use and other Permissions. For more information, send email to: permissions@pupress.princeton.edu

Follow links Class Use and other Permissions. For more information, send email to: permissions@pupress.princeton.edu COPYRIGHT NOTICE: David A. Kendrick, P. Ruben Mercado, and Hans M. Amman: Computational Economics is published by Princeton University Press and copyrighted, 2006, by Princeton University Press. All rights

More information

Models for Product Demand Forecasting with the Use of Judgmental Adjustments to Statistical Forecasts

Models for Product Demand Forecasting with the Use of Judgmental Adjustments to Statistical Forecasts Page 1 of 20 ISF 2008 Models for Product Demand Forecasting with the Use of Judgmental Adjustments to Statistical Forecasts Andrey Davydenko, Professor Robert Fildes a.davydenko@lancaster.ac.uk Lancaster

More information

Unit 31 A Hypothesis Test about Correlation and Slope in a Simple Linear Regression

Unit 31 A Hypothesis Test about Correlation and Slope in a Simple Linear Regression Unit 31 A Hypothesis Test about Correlation and Slope in a Simple Linear Regression Objectives: To perform a hypothesis test concerning the slope of a least squares line To recognize that testing for a

More information

Generalized modified linear systematic sampling scheme for finite populations

Generalized modified linear systematic sampling scheme for finite populations Hacettepe Journal of Mathematics and Statistics Volume 43 (3) (204), 529 542 Generalized modified linear systematic sampling scheme for finite populations J Subramani and Sat N Gupta Abstract The present

More information

Forecasting Methods. What is forecasting? Why is forecasting important? How can we evaluate a future demand? How do we make mistakes?

Forecasting Methods. What is forecasting? Why is forecasting important? How can we evaluate a future demand? How do we make mistakes? Forecasting Methods What is forecasting? Why is forecasting important? How can we evaluate a future demand? How do we make mistakes? Prod - Forecasting Methods Contents. FRAMEWORK OF PLANNING DECISIONS....

More information

From the help desk: Bootstrapped standard errors

From the help desk: Bootstrapped standard errors The Stata Journal (2003) 3, Number 1, pp. 71 80 From the help desk: Bootstrapped standard errors Weihua Guan Stata Corporation Abstract. Bootstrapping is a nonparametric approach for evaluating the distribution

More information

Warsaw, Poland. Abstractt. 1. Introduction. study was repeatedly data, then. calculates respect to. significant the solution.

Warsaw, Poland. Abstractt. 1. Introduction. study was repeatedly data, then. calculates respect to. significant the solution. Reports on Geodesy and Geoinformatics vol. 101/ /016; pp. 70-81 DOI: 10.1515/rgg-016-003 COMPARISON OF ROBUST ESTIMATORSS FOR LEVELING NETWORKS IN MONTE CARLO SIMULATIONS Abstractt Maria Pokarowska Faculty

More information

Premaster Statistics Tutorial 4 Full solutions

Premaster Statistics Tutorial 4 Full solutions Premaster Statistics Tutorial 4 Full solutions Regression analysis Q1 (based on Doane & Seward, 4/E, 12.7) a. Interpret the slope of the fitted regression = 125,000 + 150. b. What is the prediction for

More information