G.E.P. Box, W. Hunter, and J. Hunter. Statistics for Experimenters. Wiley, Inc., New York, 1978.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....be changed to implement equation 2.1. For example, as seen earlier, without a mask, segment 4 of 20.9.17.8 would be represented as sa4[7] But with a mask such as 0.0.0. 255, it would be, 8j255 = sa4j255] hf ; f ; f ; t; f ; f ; f ijht; t; t; ti = hsa4[7] sa4[6]; sa4[5] sa4[0]ijht; t; t; ti] ht; t; t; ti = ht; t; t; ti] t In other words, the value of this segment of the packet address is ignored, as the condition for this part of the boolean expression will always be met. After completion of the above procedures, ....

....the regression model and y is the mean of all the measured y i values. The value S y:x is roughly the average distance that the Y values are from those estimated by the model. The units of measurement of S y:x are same as the units of the dependent variable. In general, S y:x , is calculated as [6], S y:x = s P n p where y i and y i have the same meaning as above, n is the number of data points, and p is the number of model parameters estimated . The lower the value of S y:x , the better. 4.2.3 Limitations The procedure in this experiment violates one of the assumptions of ....

George E. P. Box, William G. Hunter, and J. Stuart Hunter. Statistics for Experimenters, chapter 14. John Wiley, New York, 1978.

....been pre computed for all documents. Second, the term frequencies are count data, and while they don t exactly follow the poisson distribution, it is probably the best first order approximation. The square root is well known as the variance stabilizing transformation for the poisson distribution [6], and equal variance is a desirable property for term weights. The IDF performs slightly better than the entropy measure, and using the square of the factors is marginally preferred. The practical signficance of these results is limited. It might seem that since the entropy weights use more ....

G.E.P. Box, W.G. Hunter, and J.S. Hunter. Statistics for Experimenters, page 144. John Wiley and Sons, 1978.

....2.4 The BoxCox rule. A general approachtocurve fitting is to find transformations on Y or on X, or both, that produce a straight line in the transformed scale. For example, if Y = X , then a plot of X vs p Y would produce a straight line, as would a plot of X vs Y . The Box Cox ( 1] [5]) transformation on Y is parameterized by . This transformation is applied together with a straightness statistic that permits comparisons across different parameter levels. The transformation is as follows: 8 : 1 Y ;1 if 6=0 Y ln(Y ) if =0 where Y is the geometric mean ....

....the type of bound claimed. When f (x) FPW , Y = f (X) k 2, and the NextGuess oracle includes f (x) this rule finds the function exactly. With standard normality assumptions about an added random error term, it is possible to calculate confidence intervals for the estimate b#see[1]or[5] for details. 2.5 The Difference Rule. The Difference heuristic extends Newton s divided difference method for polynomial interpolation (see [15] for an introduction) to be defined when Y contains random noise and nonpolynomial terms. The method iterates numerical differentiation on X and Y ....

G. P.Box, W. G. Hunter, and J. S. Hunter (1978) Statistics for Experimenters, Wiley & Sons.

....2.4 The BoxCox rule. A general approachtocurve fitting is to find transformations on Y or on X, or both, that produce a straight line in the transformed scale. For example, if Y = X , then a plot of X vs p Y would produce a straight line, as would a plot of X vs Y . The Box Cox ( 1] [5]) transformation on Y is parameterized by . This transformation is applied together with a straightness statistic that permits comparisons across different parameter levels. The transformation is as follows: 8 : 1 Y ;1 if 6=0 Y ln(Y ) if =0 where Y is the geometric ....

....the type of bound claimed. When f (x) FPW , Y = f (X) k 2, and the NextGuess oracle includes f (x) this rule finds the function exactly. With standard normality assumptions about an added random error term, it is possible to calculate confidence intervals for the estimate b#see[1]or[5] for details. 2.5 The Difference Rule. The Difference heuristic extends Newton s divided difference method for polynomial interpolation (see [15] for an introduction) to be defined when Y contains random noise and nonpolynomial terms. The method iterates numerical differentiation on X and Y ....

G. P.Box, W. G. Hunter, and J. S. Hunter (1978) Statistics for Experimenters, Wiley & Sons.

....yield. In practice factories try to minimize rework, so the factors identified as important to yield learning always refer to the basic process, not the rework process. B. Excursion Elimination Control charts were used to separate the chronic defects from the excursions using SQC techniques [9]. The frequency and distribution of excursions was then computed so that they could later be added to the chronic yield to compute the excursion yield. We also considered using the defect codes recorded during manufacturing to separate chronic defects and excursions. However we found that this ....

G. E. P. Box, Statistics for Experimenters, 1994.

....are in the range of 0.5 15 kJ cm . Average values for interaction time and powder density are respectively 0.5 s and 15 kW cm . Experimental design The factorial design method is one of the methods that can be used for a systematically performed parameter evaluation. It was introduced by Box [1978] and Johnson [1977] It is a multi parameter analysis method which is most useful for applications in which a process depends on numerous parameters. In such applications it is hardly possible to draw conclusions from graphs and tables that are achieved with ordinary one at a time experiments, ....

....in order to obtain the same level of accuracy. The factorial analysis gives a statistical significance to the observed relationships, which not only allows the withdrawal of insignificant process parameters from further experiments, but also allows a sensitivity analysis of the process [Box, 1978]. The calculated relationships between factors are only valid between the lower and the higher level of the parameters. Many processes in nature show an exponential course, which can not be revealed by this method. However, in a small interval such a curve can be approximated by a straight line ....

Box, G.E.P., Hunter, W.G., Hunter, J.S., 1978, Statistics for Experimenters, John Wiley & Sons, New York

....effect is a more meaningful measure of the relative importance of that factor. The sign of the effect is, of course, still needed to know if a factor treatment 5 degrades or improves the observed response. Two common methods for quickly calculating effects are the Yates algorithm (cf. 9] [11]) and the table of contrast coefficients. The next section focus on the latter, which is more intuitive and applicable to the use of Hadamard matrices. 3.2 Calculating Effects Using a Table of Contrast Coefficients Once factors of interest have been selected, systematic patterns of treatment ....

....error and to interpret the results to discard less important factors, thus focusing future experiments on those factors shown to be important. Note that the theory of Hadamard matrices [1,10,13,14,15,16] has not been covered herein. More conventional designs are found readily in numerous textbooks [7,8,9,11,12]. 11 6. ....

G.E.P Box, W.G.Hunter, J.S.Hunter, "Statistics for Experimenters," John Wiley & Sons, New York, 1978.

....the individual items in the surveys were analyzed and compared. This gave a flavor for changes occurring between the pre and post intervention surveys. The findings are given in section 7.1. Second, the difference in mean scores on outcome variables was analyzed using a paired sample t test [3]. This test compared identical measures in both surveys. The findings are presented in section 7.2. Third, correlation analysis was used to determine whether variables were associated with one another. These findings are discussed in section 7.3. 7.1 Time 1 and Time 2 Survey Findings [16 (Sec. ....

Box, G., W. Hunter, and J. Hunter. Statistics for Experimenters. John Wiley & Sons, Inc, New York, 1978.

....all transition states into account, resulting in low accuracy. Dynamic timing models, on the other hand, require the consideration of all possible input combinations and therefore suffer from the dimensionality problem. We propose to use a set of techniques known as design of experiments (DOE) [4] and response surface methods [5] to effectively deal with the optimisation problem while still being able to carry out the analysis using dynamic timing. The DOE techniques, particularly fractional factorial methods [4] can greatly reduce the number of combinations of design variables that need ....

....We propose to use a set of techniques known as design of experiments (DOE) 4] and response surface methods [5] to effectively deal with the optimisation problem while still being able to carry out the analysis using dynamic timing. The DOE techniques, particularly fractional factorial methods [4], can greatly reduce the number of combinations of design variables that need be tried (thereby ameliorating the dimensionality problem) For this given reduced set DOE, techniques greatly increase the information per simulation run [6] Response surface models (RSMs) effectively replace the slow ....

[Article contains additional citation context not shown here]

BOX, C.E.P., HUNTER, W.G., HUNTER, J.S.: `Statistics for experimenters' (John Wiley & Sons, 1987)

....the information gained from the complete set of genes in estimated joint prior distributions for ( i )and(# 2 i ) We let the prior distribution of 1 # 2 i be gamma, and that of i given # 2 i be normal. This is a conjugate prior distribution allowing us to calculate B i explicitly. See e.g. [1], 9] 2] and [6] For # degrees of freedom and scale parameters a 0, c 0, we set # i = na 2# 2 i and suppose that # i # #(#, 1) 6a) i # i # =0 ifI i =0 N(0,cna 2# i )ifI i =1 (6b) for all i =1. N. The parameter c expressing dependence between the priors for ( i ) and (# i ) is ....

....0.001. The choice of p does not change the shape of the B vs M . plot very much, but it does move it up and down the y axis. A consequence of this is that we cannot fix a cuto# (e.g. B = 0, as suggested for (5) to select all the genes with a higher score as di#erentially expressed. References [1] J Aitchison and I R Dunsmore. Statistical prediction analysis. Cambribge University Press, 1975. 2] G E P Box and G C Tiao. Bayesian inference in statistical analysis. Addison Wesley, 1973. 3] M J Buckley. The Spot user s guide. CSIRO Mathematical and Information Sciences, 2000 Aug. ....

[Article contains additional citation context not shown here]

G E P Box, W G Hunter, and J S Hunter. Statistics for experimenters. Wiley Interscience, 1978. REFERENCES 25

....single relevant document. We can examine the data more closely to see how badly the assumption of equal variance is violated. Let us presume that the variance is constant and equal to the error variance obtained from the ANOVA model, which we will denote as oe 2 . It is well known in statistics [1] that: n Gamma 1)s 2 oe 2 2 n Gamma1 where s 2 is the variance of a sample of size n Gamma 1 drawn from a normal population with variance oe 2 . We can compute s 2 for each query and compare the distribution of the test statistic above to the appropriate chi square reference ....

G.E.P. Box, W.G. Hunter, and J.S. Hunter. Statistics for Experimenters, pages 118--119. John Wiley and Sons, 1978.

....medical statistics. The CONSORT statement on reporting the results of randomized trials in medicine [4] This statement has been adopted by seventy medical journals. 3 . A paper defining guidelines for reporting results of statistical analysis in the Japanese Journal of Clinical Oncology [12]. This paper also discusses common statistical errors. A paper defining guidelines for the American Psychological Society [49] Three papers discussing the types of statistical errors found in medical papers [31] 2] 38] We have concentrated on medical guidelines because medical ....

....P1: Describe or cite a reference for all statistical procedures used. Following from the analysis guidelines it is important to document all statistical procedures. Reference to a statistical package is not sufficient. However, there are some exceptions to that rule. For example Fukuda and Ohashi [12] suggest that the following statistics do not require explicit references: t test, simple chi squared test, Wilcoxon or Mann Whitney U test, correlation and linear regression. P2: Report the statistical package used Statistical packages often give slightly different results. So, it is important ....

[Article contains additional citation context not shown here]

G. E. P. Box, W. G. Hunter and J. S. Hunter, Statistics for Experimenters, John Wiley and Sons, New York, 1978.

....Analysis of Variance dialog opens, as shown in Figure 8.14. Example The simplest kind of experiments are those in which a single continuous response variable is measured a number of times for each of several levels of some experimental factor. For example, consider the data in Table 8. 2 (from Box, Hunter, and Hunter (1978)) The data consist of numerical values of blood coagulation times for each of four diets. Coagulation time is the continuous response variable, and diet is a qualitative variable, or factor, having four levels: A, B, C, and D. The diets corresponding to the levels A, B, C, and D were determined ....

....different in at least one of the groups. Performing a Friedman rank test From the main menu, choose Statistics c Compare Samples c k Samples c Friedman Rank Test. The Friedman Rank Sum Test dialog opens, as shown in Figure 8.17. Example The data set shown in Table 8. 3 was first used by Box, Hunter, and Hunter in 1978. The data was collected to determine the effect of treatments A, B, C, and D on the yield of penicillin in a penicillin manufacturing process. The response variable is yield, and the Figure 8.17: The Friedman Rank Sum Test dialog. y y Chapter 8 Statistics 306 treatment variable is treatment. ....

Box, G.E.P., Hunter, W.G., & Hunter, J.S. (1978). Statistics for Experimenters. New York: Wiley.

....# easy to implement. # Usage of the assignment as calculated by the linear optimization or the practical #rule of thumb# one. These four parameters are analysed using a 2 k factorial design, giving 16 design points #for description of experimental design, see #Law and Kelton, 1992#, #Box et al. 1978##. The coding chart is shown in table 2. Factor j Parameter 1 N p 2 4 2 # p 45 55 3 # r 35 45 4 assignment thumb LP Table 2: Coding chart For each design point, 5 replications are made, the results are shown in #gure 2. A steady state is reached after a simulation run time of 2h, every ....

Box, G. E., Hunter, W. G., and Hunter, J. S. #1978#. Statistics for Experimenters. John Wiley & Sons.

....and the long tailedness of the data, another clue that to my knowledge has not yet been pursued further. Yet these studies (and a few others) were completely ignored by the mainstream of statistics. People continued (and mostly still continue) to believe in the declaration of independence (Box et al. 1978) instead of hunting out the real uncertainty and noticing how oe= p n can mislead (Mosteller and Tukey 1977) This attitude dominates statistics virtually uncontested, although also in our times, top applied statisticians like Cuthbert Daniel (cf. Daniel and Wood 1980, and the two books just ....

Box, G. E. P., Hunter, W. G. and Hunter, J. S. (1978). Statistics for Experimenters, Wiley, New York.

.... correlated with . Suppose, for example, the defining relation of a factorial design is I=ABCF and the splitting generators are: B 1 ABE; B 2 BCDE. Since there are no expected block effects, these effects are eliminated. The words ABE and BCDE are then used to extend the defining relation (see Box, Hunter and Hunter, 1978, page 409) to a correlation relation as follows: I=ABCF ABE CEF BCDE ADEF ACD BDF. Multiplying any effect by this correlation relation shows the confounding and correlation pattern. Concepts similar to resolution and aberration (see Fries and Hunter, 1980) can now be used to select splitting ....

Box, G. E. P., Hunter, W. G., and Hunter, J. S. (1978), Statistics for Experimenters, New York: John Wiley.

....resulting in costefficiency. We prefer a systematic approach and use Design of Experiments (DOE) principles to conduct a minimum ordered set of numerical experiments to generate response surfaces. The DOE techniques are formal techniques which support the design and analysis of experiments (Box, et al. 1978; Montgomery, 1991) Among various DOE techniques, the Response Surface Methodology (RSM) Box and Draper, 1987; Khuri and Cornell, 1987) is a collection of statistical techniques which support the design of experiments and fitting response surface models. These response surface models are ....

....experiments conducted and the accuracy of the response model. Since computer experiments differ from physical experiments in that there is no random error (Sacks, et al. 1989; Welch, et al. 1990) classical methods for the design and analysis of physical experiments (e. g, Box and Draper, 1987; Box, et al. 1978) are no long ideal for complex, deterministic computer models. In this paper, we present a sequential experimentation strategy as well as the heuristic rules for creating high order response surface models as a cost effective approach to applying statistical experimentation methods in design ....

[Article contains additional citation context not shown here]

Box, G. E. P., Hunter, W. G. and Hunter, J. S., 1978, Statistics for Experimenters, John Wiley & Sons, New York.

.... this study (Senkandwa 1998) In order to reduce the waiting time at the DMV, simulation was chosen as the method of analysis because it easily lends itself to incorporating the complexity of the relationships between system components (Banks et al. 1996) Furthermore, a 2 3 factorial experiment (Box et al. 1978) was performed to test the sensitivity of time spent at the DMV to the following factors: interarrival time, number of servers, and service time. 2 SYSTEM DESCRIPTION Figure 1 illustrates the DMV system. After entering the DMV system, customers may seek help at the information desk. While they ....

Box, G., Hunter, J., Hunter, W. 1978. Statistics For Experimenters. Canada: John Wiley & Sons, Inc.

....the design simulation, while the Aircraft Life Cycle Cost Analysis (ALCCA) 10] program was selected to be used instead of the economics package within FLOPS to allow for an estimation of such economic output variables as RPM, DOC, and TOC. After selecting the synthesis tool, a screening test [11] that monitors the linear sensitivities for all pertinent variables is conducted to identify the most important variables through the use of the Pareto principle [12] Based on previous studies performed by the authors [1] 2] 3] an inclusive list of design and economic variables were ....

....Design of Experiment [4] table is set up which defines combinations of the nine variables determined in Table II. The set of these combinations describe the design space while each individual combination represents a separate execution of the synthesis code. An analysis of variance (ANOVA) [11] performed on the DOE response values determines the functional relationship of response ( RPM, DOC, TOC) and the control and noise variables. For this study, the typically assumed second order regression model [4] 11] is extended with higher order interaction terms to estimate the influence ....

[Article contains additional citation context not shown here]

Box, G.E.P., Hunter, W.G., Hunter, J.S., Statistics for Experimenters, John Wiley & Sons, Inc., New York, NY, 1978.

.... correlated with . Suppose, for example, the defining relation of a factorial design is I=ABCF and the splitting generators are: B 1 ABE; B 2 BCDE. Since there are no expected block effects, these effects are eliminated. The words ABE and BCDE are then used to extend the defining relation (see Box, Hunter and Hunter, 1978, page 409) to a correlation relation as follows: I=ABCF ABE CEF BCDE ADEF ACD BDF. Multiplying any effect by this correlation relation shows the confounding and correlation pattern. Concepts similar to resolution and aberration (see Fries and Hunter, 1980) can now be used to select splitting ....

Box, G. E. P., Hunter, W. G., and Hunter, J. S. (1978), Statistics for Experimenters, New York: John Wiley.

....the other, less important input parameters. The second problem with Method #2 is in the mathematical background of such regression methods as Response Surface Methodology (RSM) and Design Of Experiments (DOE) which are based on random rather than deterministic variables (see [10] 12] 22] [23]) Therefore, the authors would like to stress here the use of caution in the straightforward application of these methods. Fundamental statistical knowledge is critical for obtaining reasonable approximations of the computer model. Many of the statistical analysis results the commercial packages ....

....design methods. 8] In order to compare that approach (Method #2) with the one introduced in this paper (Method #3) a very brief overview of the combined Response Surface Equation Monte Carlo Simulation (RSE MCS) method is provided here. For more detailed information refer to [6] 8] 22] and [23]. RSM is based on a statistical approach to build and rapidly assess empirical metamodels. 22, 23] By carefully designing and analyzing experiments or simulations, the methodology seeks to relate and identify the relative contributions of various input variables to the system response. However, ....

[Article contains additional citation context not shown here]

Box, G.E.P., Hunter, W.G., Hunter, J.S., Statistics for Experimenters, John Wiley & Sons, Inc., New York, NY, 1978 11

....different 95 confidence interval. Any one of these intervals has a 95 a priori probability of containing the true difference for the two systems in the pair. However, the probability that all k(k 1) 2 intervals will simultaneously include their true parameter values is somewhat less. The text by Box, Hunter, and Hunter [1978] describes several statistical techniques developed to address the problem of multiple comparisons. Often, no special action is needed as long as the investigator is aware of the issue when interpreting the data. Another consideration involves controlled experimental factors that are expected to ....

....outlined in this report are to be conducted separately for each combination of factors tested. In addition, the statistical method known as 22 the analysis of variance (ANOVA) can be used to determine which factors (mine size, mine type, burial depth, soil moisture, etc. impact performance [Box, Hunter, and Hunter, 1978]. For many mine detection systems, the primary output is not a binary ( mine no mine ) decision, but rather a continuous variable, expressed as a function of location (or time) that is thresholded to determine where (or when) detections have occurred. In this report, only the binary results have ....

Box, G.E.P, Hunter, W.G., and Hunter, J.S. (1978). Statistics for Experimenters. New York: John Wiley & Sons.

....good estimates of the effects that are deemed the most important. The general field of statistical experimental design is concerned with the planning of efficient experiments to provide the desired information in a readily extractable form. Two standard references in this area are the texts by Box, Hunter, and Hunter (1978) and Cox (1958) The first step in analyzing data from multi factor experiments often involves using the analysis of variance to determine which factors (and combinations of factors) significantly impact performance. While ANOVA is a familiar and widely used technique, it is based on several ....

....the normal approximation may be inadequate if the observed percentage is close to zero or one, or if the number of target mines emplaced is small. In such cases, it is necessary to transform the observed percentages to a new domain in which the normality assumption is more nearly met. The book by Box, Hunter, and Hunter (1978) provides an excellent introduction to the mathematics of ANOVA. The collection edited by Hoaglin, Mosteller, and Tukey (1991) discusses a number of more advanced topics, including transformation and graphical display. It is hoped that the material covered in the present report and its companion, ....

Box, G.E.P., Hunter, W.G., and Hunter, J.S. (1978). Statistics for Experimenters. New York: John Wiley & Sons.

....until a leaf was encountered. The percentage of objects correctly identified was then determined by totaling the matches between the instance classes (positive or negative) at the leaves and those of the test objects. The analysis of variance model was a 2 4 x5x15 full factorial design [CC50] [BHH78]. Since the tree generation and analysis was automated, the increased sensitivity of a full factorial design was preferred over a more economical fractionalfactorial design[Sch59] There were six independent variables and four dependent variables. The data analysis included the 15 two way ....

G. E. P. Box, W. G. Hunter, and J. S. Hunter. Statistics for Experimenters. John Wiley & Sons, New York, 1978.

....cost, and data availability constraints. The classification tree generation parameters can be calibrated to enhance performance in a given environment. In order to assess tree effectiveness with a variety of tree generation parameters, an analysis of variance statistical model was applied [CC50] BHH78] The model required 12,000 trees to be generated and analyzed. The analysis of variance model was a 2 2 Theta 5 2 Theta 15 full factorial design. There were 5 independent variables and 6 dependent variables. The data analysis included all two way and three way interactions among the five ....

G. E. P. Box, W. G. Hunter, and J. S. Hunter. Statistics for Experimenters. John Wiley & Sons, New York, 1978.

....in jaggies. 11 Figure 1: Plot of Equation 1 for 1 x 2 . The default for spin plot is white on black. If you wish to print one of these, you should change it (using the right menu button) to black on white. Otherwise you ll get a page of black ink. 13 Function plots Plots like those in Box et al. 1978, pp. 296 and 297) are quite easy to produce in XLISPSTAT. For example, to plot the function f(x) 5 2x x 2 (1) over the range 1 x 2, we can use the XLISPSTAT commands (defun f (x) 5 ( 2 x) 1 ( x 2) plot function # f 1 2) The first line defines the function f and the ....

....N standard normal random numbers. Vectorized. NIL 16 One way analysis of variance The XLISPSTAT oneway model function is used to perform a one way ANOVA analysis. Below is an example session log for the blood coagulation data in Box et al. 1978, Table 6.1, p. 166) Some points to observe as you compare the results here with those in the text: The results from oneway model are saved in an object called anova (the name is arbitrary) This allows the residuals and fitted values to be retrieved by sending messages to the anova object. ....

[Article contains additional citation context not shown here]

G. E. P. Box, W. G. Hunter, and J. S. Hunter. Statistics for Experimenters. Wiley, New York, 1978.

....steps one more time. Analysis Strategy The goal of this analysis is to determine whether the data supports our hypothesis that the interaction of process environment and task priorities explain variation in inspection interval. We analyzed inspection interval using linear regression models [2]. We built one model for the pre meeting interval (time from availability to meeting) and another for the post meeting interval (time from meeting to the completion of repair) Our reasoning was that since only one reviewer (the moderator) is involved after the meeting, it is likely that different ....

George E. Box, William G. Hunter, and J. Stuart Hunter. Statistics for Experimenters. John Wiley and Sons, Inc., 1978.

....needed to achieve a feasible task set. To simplify simulation, we set the rejection penalty = 0, and assume that all transactions are ATAR. This allows us to negotiate without the need to account for the constraints shown in the upper part of Figure 11. First, we make a 2 level factorial [40] design to determine important parameters. We test the following parameters, E(nrtproblems) with values 2 and 8, E(ntransactions) with values 2 and 8, E(ntasks) with values 2 and 8, E(wcet) with values 20 and 40, E(nqoses) with values 2 and 8 and E(scaledeadlineperiod) with values 50 and 100. In ....

G. E. P. Box, W. G. Hunter, and J. S. Hunter. Statistics for Experimenters, chapter 10, page 653. John Wiley and Sons, 1978. 45

....experiment, it is very challenging to make useful inference when some observations are censored. Traditionally, simple and straightforward methods are used for analyzing data from experimental designs, such as the Yates algorithm for algebraic calculation of main effects and interactions (see Box, Hunter, and Hunter 1978), half normal plot method (Daniel 1959) and linear regression models. However, these methods are not adequate when censored observations are obtained from the experiments. To deal with this diculty, several methods have been proposed in the literature. Hahn, Morgan, and Schmee (1981) developed an ....

Box, G.E.P, Hunter, W.G., and Hunter, J.S. (1978), Statistics for Experimenters, New York: John Wiley.

....execution has completed. S Check [50,71] a tool developed at NIST, uses a partially automated approach to diagnose performance critical parts of large scale applications. The tool uses artificially introduced delays together with the statistical technique known as Design of Experiments or DEX [4] to focus attention on synchronization points that are the cause or potential cause of performance bottlenecks. In their approach, called Synthetic Perturbation Screening, parts of the experimental setup, such as narrowing down potentially fruitful locations to test, are done manually through a ....

....8. This match function uses resource string labels as the basis for determining resource difference. We discuss resource matching further in Section 3.4.5. EventMap E 1 , E 2 ) returns EventMap [1] E [2] r i , T i ) E 1 [3] if (r j , T j ) E 2 , s.t. match (r i , r j ) then [4] E E r i r j , T i T j [5] E 2 E 2 (r j , T j ) 6] else E E U (r i , T i ) 7] E E E 2 [8] return E Figure 6: Algorithm to find the Structural Merge of two EventMaps, E 1 E 2 34 The structural merge operation has the following properties: commutativity: E 1 E 2 = E ....

[Article contains additional citation context not shown here]

G.E.P. Box, W.G. Hunter, and J.S. Hunter. Statistics for Experimenters. John Wiley and Sons, New York, 1978.

....which is a problem generator developed by Kolisch and Sprecher (1997) This problem set is specifically designed for the RCPSP, and it consists of 480 instances with n = 30 activities and K = 4 resources. The weights # i (i = 1, n) were randomly drawn from the discrete uniform distribution U[1, 5]. 4.3 Parameters selection Our MOTS algorithm allows for di#erent choices of five parameters: the tabu list size TL, the number of iterations before stopping Iter max , the number of steps s, the neighborhood size N , and the aggregation function. We performed some preliminary experiments and we ....

Box G., W. Hunter, S. Hunter (1978), Statistics for experimenters, John Wiley & Sons.

No context found.

G.E.P. Box, W. Hunter, and J. Hunter. Statistics for Experimenters. Wiley, Inc., New York, 1978.

No context found.

G. E. P. Box, W. G. Hunter, and J. S. Hunter. Statistics for Experimenters. John Wiley and Sons, Inc., 1978.

No context found.

G. E. P. Box, W. G. Hunter, and J. S. Hunter. Statistics for Experimenters. John Wiley, 1978.

No context found.

Box, G E P, Hunter, W G , and Hunter, J S ,1978, Statistics for Experimenters, John Wiley & Sons.

No context found.

G. E. P. Box, W. G. Hunter, and J. S. Hunter. Statistics for Experimenters. John Wiley, 1978.

No context found.

G. E. P. Box, W. G. Hunter, and J. S. Hunter. Statistics for Experimenters. John Wiley, 1978.

No context found.

B. E. P. Box, W. G. Hunter and J. S. Hunter, Statistics for Experimenters, John Wiley & Sons, New York, 1978.

No context found.

George E. P. Box, William G. Hunter and J. Stuart Hunter. Statistics for Experimenters. John Wiley & Sons, 1978.

No context found.

Box, G.E.P., Hunter, W.G., Hunter, J.S., Statistics for Experimenters. John Wiley & Sons, Inc., New York, 1978.

No context found.

G.E.P. Bock, W.G. Hunter, and J. Hunter. Statistics for Experimenters. John Wiley & Sons, New York, 1978.

No context found.

) Box, G. E. P., Hunter, W. G., and Hunter, J. S. (1978). Statistics for Experimenters. An Introduction to Design, Data Analysis, and Model Building. New York: Wiley.

No context found.

Box, G.E.P., Hunter, W.G., Hunter, J.S., Statistics for Experimenters, John Wiley & Sons, Inc., New York, 1978.

No context found.

Box, G.E.P., E.P. Hunter and J.S. Hunter (1978). Statistics for Experimenters, Wiley, New York.

No context found.

) Box, G.E.P., Hunter, W.G., Hunter, J.S., Statistics for Experimenters, John Wiley & Sons,

No context found.

Box, G.E.P., Hunter, W.G., Hunter, J.S., Statistics for Experimenters, John Wiley & Sons, Inc., New York, NY, 1978.

No context found.

Box, G. E. P., W. G. Hunter, et al. (1978). Statistics for Experimenters. New York, John Wiley and Sons.

No context found.

Box, G.E.P., Hunter, W.G. and Hunter, J.S., Statistics for Experimenters, John Wiley and Sons, New York, 1978.

No context found.

Technometrics 36, pp. 308-311. Box, G. E. P.; Hunter, W. G.; and Hunter, J. S. (1978). Statistics for Experimenters. John Wiley, New York, NY.

No context found.

George E. Box, William G. Hunter, and J. Stuart Hunter. Statistics for Experimenters. John Wiley and Sons, Inc., 1978.

First 50 documents  Next 50

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC