a b s t r a c t The interest in nonparametric statistical analysis has grown recently in the field of computational intelligence. In many experimental studies, the lack of the required properties for a proper application of parametric procedures - independence, normality, and homoscedasticity - yields to nonparametric ones the task of performing a rigorous comparison among algorithms. In this paper, we will discuss the basics and give a survey of a complete set of nonparametric procedures developed to perform both pairwise and multiple comparisons, for multi-problem analysis. The test problems of the CEC'2005 special session on real parameter optimization will help to illustrate the use of the tests throughout this tutorial, analyzing the results of a set of well-known evolutionary and swarm intelligence algorithms. This tutorial is concluded with a compilation of considerations and recommendations, which will guide practitioners when using these tests to contrast their experimental results.

https://sci2s.ugr.es/sites/default/files/ficherosPublicaciones/1374_2011-Derrac-SWEVO.pdf

A practical tutorial on the use of nonparametric statistical tests as a methodology for comparing evolutionary and swarm intelligence algorithms

https://sci2s.ugr.es/sites/default/files/files/TematicWebSites/sicidm/2010-Garcia-INS.pdf

Advanced nonparametric tests for multiple comparisons in the design of experiments in computational intelligence and data mining: Experimental analysis of power

The solution of the segmentation problem requires a mechanism for partitioning the image array into low-level entities based on a model of the underlying image structure. A piecewise-smooth surface model for image data that possesses surface coherence properties is used to develop an algorithm that simultaneously segments a large class of images into regions of arbitrary shape and approximates image data with bivariate functions so that it is possible to compute a complete, noiseless image reconstruction based on the extracted functions and regions. Surface curvature sign labeling provides an initial coarse image segmentation, which is refined by an iterative region-growing method based on variable-order surface fitting. Experimental results show the algorithm's performance on six range images and three intensity images. >

/pdf/segmentation-through-variable-order-surface-fitting-29h4kwrbaf.pdf

Segmentation through variable-order surface fitting

Abdominal obesity has been suggested to be associated with perturbations of the regulation of the hypothalamic-pituitary-adrenal (HPA) axis. In a population of 51-yr-old men (n = 284) salivary cortisol concentrations were determined on repeated (n = 7) occasions over a random working day, and perceived stress was reported in parallel. Cortisol values were then related to reported stress (stress-related cortisol). A standardized lunch was used as a physiological challenge. A low dose (0.5 mg) dexamethasone suppression test was also performed as well as determinations of testosterone and insulin-like growth factor I (IGF-I). Body mass index [weight (kilograms)/height (meters)2]; waist/hip circumference ratio (WHR); sagittal trunk recumbent diameter (D); fasting insulin; blood glucose; triglycerides; and total, low density (LDL), and high density (HDL) lipoprotein cholesterol were also determined. Cortisol concentrations were highest in the morning, and lunch was followed by a peak (P = 0.044). Two types of diurnal cortisol curves were identified, one characterized by a high variability with high morning values, and another with low variability and low morning values. Both correlated strongly with suppression of salivary cortisol by dexamethasone (P < 0.001). Stress-related cortisol secretion was associated with D (P = 0.051), low IGF-I (P = 0.006), and diastolic blood pressure (P = 0.078). When the type of diurnal cortisol curve was taken into consideration by statistical weighting, stress-related cortisol secretion in subjects with high variability showed associations with testosterone (P < 0.001), D, total and LDL cholesterol, diastolic blood pressure (P < 0.001), fasting insulin (P = 0.039), and glucose (P = 0.030) as well as, negatively, triglycerides (P < 0.001). When weighted for a low variability of diurnal cortisol secretion, stress-related cortisol secretion showed strong negative relationships with IGF-I, testosterone, and HDL. Furthermore, strong, consistent relationships (all P < 0.001) were found with obesity factors (body mass index, WHR, and D), and with metabolic (insulin, glucose, triglycerides, and total and LDL cholesterol) as well as hemodynamic variables (systolic and diastolic blood pressure and heart rate). These results clearly show interactions between diurnal cortisol secretion related to perceived stress and anthropometric, endocrine, metabolic, and hemodynamic variables. This seems to occur with apparently normal regulation of the HPA axis (high morning peaks and variability as well as dexamethasone suppression of cortisol), where other endocrine variables are not affected. With a low diurnal cortisol variation and blunted dexamethasone suppression, indicating abnormal regulation of the HPA axis, perceived stress-dependent cortisol values were strongly related to perturbations of other endocrine axes as well as abdominal obesity with metabolic and hemodynamic abnormalities. Perturbations of the regulations of the HPA axis such as those described in combination with low dexamethasone suppressibility are known to follow long term overactivation of the axis by factors such as environmental stress.

/pdf/stress-related-cortisol-secretion-in-men-relationships-with-4j4mgmkprm.pdf

Stress-related cortisol secretion in men: relationships with abdominal obesity and endocrine, metabolic and hemodynamic abnormalities

K-means is undoubtedly the most widely used partitional clustering algorithm. Unfortunately, due to its gradient descent nature, this algorithm is highly sensitive to the initial placement of the cluster centers. Numerous initialization methods have been proposed to address this problem. In this paper, we first present an overview of these methods with an emphasis on their computational efficiency. We then compare eight commonly used linear time complexity initialization methods on a large and diverse collection of data sets using various performance criteria. Finally, we analyze the experimental results using non-parametric statistical tests and provide recommendations for practitioners. We demonstrate that popular initialization methods often perform poorly and that there are in fact strong alternatives to these methods.

A comparative study of efficient initialization methods for the k-means clustering algorithm

Cengage Learning, 2000. Book Condition: New. Brand New, Unread Copy in Perfect Condition. A+ Customer Service! Summary:

Nathan Jaspen

Papers

Applied Nonparametric Statistics