Firework Plots for Evaluating the lmpact of Outliers and lnfluential Observations in Generalized Linear Models

( Dae Heung Jang ),( Cahristinem. Anderson Cook ) 한국품질경영학회 2014 한국품질경영학회 학술대회 Vol.2014 No.2

Outliers can distort many measures in data analysis and statistical modeling, and influential points can have disproportionate impact on the estimated values of model parameters. Jang and Anderson-Cook [1] proposed a new set of graphical summaries, called firework plots, as simple tools for evaluating the impact of outliers and influential points in regression. Variations of the plots focus on allowing visualization of the impact on the estimated parameters and variability. In the generalized linear models analysis setting, the impact of changing model parameters is often less transparent than in the linear model setting and variability can be captured with the deviance. Hence, this paper describes how 3-D firework plots and the pairwise firework plot matrix can be used to increase understanding of contributions of individual observations and as a complement to other regression diagnostics techniques in the generalized linear models setting. Using these firework plots, we can find outliers and influential points and their impact on model parameters. We illustrate the information and understanding gain possible with several examples.

Competing Views on Law and Adjudication in Canada: The Black-Letter Law Tradition versus Contextualism

Bruce Anderson(Bruce Anderson ) 한국캐나다학회 2003 Asia-Pacific Journal of Canadian Studies (APJCS) Vol.9 No.-

The Set of Torsion Elements of a Module

Anderson, D. D.,Chun, Sangmin Taylor Francis 2014 Communications in Algebra Vol.42 No.4

AGREEABLE DOMAINS

Anderson, D. D.,Kwak, Dong Je,Zafrullah, Muhammad 경북대학교 위상수학 기하학연구센터 1995 硏究論文集 Vol.4 No.-

An integral domain D with quotient field K is defined to be agreeable if for each fractional ideal F of D[X] with F ⊆ K[X] there exists 0 ≠ s ∈ D with sF ⊆ D[X]. D is agreeable ⇔ D satisfies property (*) (for 0 ≠ f(X) ∈ K[X], there exists 0 ≠ s ∈ D so that f(X)g(X) ∈ D[X] for g(X) ∈ K(X) implies that sg(X) ∈ D[X]) ⇔ D[X] is an almost principal domain, i.e., for each nonzero ideal I of D[X] with IK[X] ≠ K[X], there exists f(X) ∈ I and 0 ≠ s ∈ D with sI ⊆ (f(X)). If D is Noetherian or integrally closed, then D is agreeable. A number of other characterizations of agreeable domains are given as are a number of stability properties. For example, if D is agreeable, so is ??Dp_(a) and for a pair of domains D ⊆ D′ with [D: D′] ≠ 0, D is agreeable ⇔ D′ is agreeable. Results on agreeable domains are used to give an alternative treatment of Querre’s characterization of divisorial ideals in integrally closed polynomial rings. Finally, the various characterizations of D being agreeable are considered for polynomial rings in several variables.

How Can South Asia and Sub-Saharan Africa Gain From the Next WTO Round?

Anderson, Kym,Yao, Shunli 세종대학교 국제경제연구소 2003 Journal of Economic Integration Vol.18 No.3

If South Asia and Sub-Saharan Africa are to become constructively engaged in the next attempt by World Trade Organization(WTO) members to liveralize trade multilaterally, they need to be convinced that there will ve sufficient gains from trade reform to warrant the inevitable costs of negotiation and adjustment. This paper provides new estinates of the likely economic effects on their evonomies of further liberalizing world trade post-Uruguay Round. The results show that the developing countries of South Asia and Sub-Sanaran Africa have much to gain from taking part in the next round. However, those gains will ve far greater the more those countries are willing to embrace regorm at home so as to enable their firms to take greatest advantage of the opportunities provided by the opening up of markets abroad.

Preparation and Characterization of Polymeric Composites Assembled from Fiberglass Fabric Waste from the Wind Blades Manufacturing Process

Anderson P. Fernandes,Karen S. Prado,Maria N. Castanho,Jane M. F. Paiva 한국섬유공학회 2022 Fibers and polymers Vol.23 No.12

Currently, there is still no knowledge of the potential uses of the increasing post-industrial fiberglass fabric wastegenerated from wind turbine manufacturing industries. The present work proposes reusing this fiberglass fabric waste asreinforcement in new thermosetting polymer composites. Polyester composites with different numbers and stackingsequences of stitched and random scraps of fiberglass fabrics were prepared using a resin transfer molding process. Moreover, their thermal stability, flexural and tensile properties were evaluated. Results of the statistical analysis performedaccording to Tukey’s methodology indicated that the number of fiberglass layers is the decisive factor influencing themechanical properties of the developed composites, whereas their stacking sequence does not have a significant influence. The composites with four layers of fiberglass fabric waste showed promising mechanical properties with flexural and tensilestrengths higher than 205 and 115 MPa, respectively, and Young’s Modulus higher than 2700 MPa, enabling their use innumerous applications. On the other hand, the number of fiberglass fabric layers had only a small influence on the thermalstability of the composites of around 335 ºC. Overall results showed that reusing fiberglass fabric waste as reinforcement innew polymer composites is a feasible way to valorize this waste from wind turbine industries.

The Optimal Borrowing Tax

Anderson, James E. 세종대학교 국제경제연구소 1991 Journal of Economic Integration Vol.6 No.1

Optimal public finance analysis of intertemporal models has been devoted to special cases where the distortion of intertemporal choice is avoided. In contrast, practical public finance has significant distortion of intertemporal choice. This paper characterizes optimal intertemporal taxation in a simple model. With distortionary taxation necessary, fluctuations of productivity of external real interest rates are potent possible reasons to distort intertemporal choice. Borrowing/lending taxation is shown to be generally required for efficiency, even in the special separable case usually thougth to imply no intertemporal distortion. The time profile of the optimal borrowing/lending tax is characterized in a special case, and the resulting relation between optimal public debt and the trade account is analyzed.

Annihilator conditions on modules over commutative rings

Anderson, D. D.,Chun, Sangmin WORLD SCIENTIFIC PUBLISHING COMPANY 2017 JOURNAL OF ALGEBRA AND ITS APPLICATIONS Vol.16 No.8

<P>Let <TEX>$ R$</TEX> be a commutative ring and <TEX>$ M$</TEX> an <TEX>$ R$</TEX>-module. Let <TEX>$ Z(M) = \{r \in R\,|\,rm = 0\text{ for some nonzero }m \in M\}$</TEX> and <TEX>$ T(M) = \{m \in M\,|\,rm = 0\text{ for some nonzero }r \in R\}$</TEX>. <TEX>$ M$</TEX> satisfies Property <TEX>$ A$</TEX> (respectively, Property <TEX>$ T$</TEX>) if for each finitely generated ideal <TEX>$ I \subseteq Z(M)$</TEX> (respectively, finitely generated submodule <TEX>$ N \subseteq T(M)$</TEX>) ann<TEX>$ _{M}(I)\neq 0$</TEX> (respectively, ann<TEX>$ _{R}(N)\neq 0$</TEX>). The ring <TEX>$ R$</TEX> satisfies Property <TEX>$ A$</TEX> if <TEX>$ _{R}R$</TEX> does. We study rings and modules satisfying Property <TEX>$ A$</TEX> or Property <TEX>$ T$</TEX>. A number of examples are given, many using the method of idealization.</P>

Strategic Lobbying and Antidumping

Anderson, James E. 세종대학교 국제경제연구소 1994 Journal of Economic Integration Vol.9 No.2

Anti-dumping is often defended as a pressure valve which reduces more illiberal forms of protectionist pressure. In the domino dumping model of Anderson [1992, 1993] this need not be true as exporters dump to obtain market access in the event of a VER. The contribution of this paper is to show that anti- dumping opens a channel for strategic lobbying through which lobbying commitment can have favorable effects on the decisions of exporting firms, and through which antidumping enforcement can encourage lobbying. Thus a 'de-politicizing' institution can perversely be responsible for politicizing trade policy all the more.

Ascophyllum Nodosum Supplementation Strategies That Improve Overall Carcass Merit of Implanted English Crossbred Cattle

Anderson, M.J.,Blanton, J.R. Jr.,Gleghorn, J.,Kim, S.W.,Johnson, J.W. Asian Australasian Association of Animal Productio 2006 Animal Bioscience Vol.19 No.10

English crossbred steers (n = 32) and heifers (n = 32) were fed a corn-based finishing diet and supplemented with 2% Ascophyllum nodosum on a DM basis to for various feeding stages of the 119 d feeding period determine the supplementation strategy that maximized intramuscular fat deposition as determined by quality grade. All cattle were implanted with Ralgro$^{(R)}$ on d 36 of the trial and re-implanted with Revalor-S$^{(R)}$ or Revalor-H$^{(R)}$ on d 92. Cattle were blocked by sex and divided into one control and three treatment groups receiving Ascophyllum nodosum. Treatment 1 (trt 1) received Ascophyllum nodosum from d 36 to 50 of the feeding period, trt 2 received Ascophyllum nodosum for the last 14 d of the feeding period, and trt 3 which received Ascophyllum nodosum for both d 36-50 and the last 14 d of the feeding period. Cattle were weighted initially ($385{\pm}4.53kg$) and every 28 d following until they reach an average BW of $554{\pm}7.46kg$. No effect for Ascophyllum nodosum supplementation was found on measured performance characteristics. All treatment groups supplemented with Ascophyllum nodosum had higher actual marbling scores (p<0.05) than controls. Trt 1 was found to have a highest marbling score (572.5; p<0.05), whereas the control group having the lowest marbling score (473.8). Trt 1 had a higher quality grade (5.25; p<0.05) than the control (3.94) group, but did not differ from trt 2 (4.56; p = 0.105) and trt 3 (4.75; p = 0.236) where high Select = 4, and low Choice = 5. Trt 2 did not differ from trt 3 (p = 0.655), or the control group (p = 0.140) for quality grade. However, trt 3 did tend to differ (p = 0.057) from the control group for quality grade. Control group animals graded 25% Choice, 62.5% Select; trt 1 graded 75% Choice, 18.8% Select; trt 2 graded 62.5% Choice, 25% Select and trt 3 graded 56.3% Choice and 31.2% Select. Overall, treatment groups had a 39.6% increase in Choice quality grade and a 37.5% decrease in Select quality grade when compared to the control animals.