Automatically determining the number of clusters without a priori knowledge is a difficult research issue for data clustering problem. An effective multiobjective evolutionary algorithm based clustering algorithm is proposed to not only overcome this problem but also provide a better clustering result in this study. The proposed algorithm differs from the traditional evolutionary algorithm in the sense that instead of a single crossover operator and a single mutation operator, the proposed algorithm uses a pool of crossover operators and a pool of mutation operators that are selected at random to increase the search diversity. To evaluate the performance of the proposed algorithm, several well-known datasets are used. The simulation results show that not only can the proposed algorithm automatically determine the number of clusters, but it can also provide a better clustering result.

The intent of this study was to investigate a system approach to implementing a diversity plan in a chosen school district, with emphasis on (1) why a chosen school district utilized a systems approach in implementing a diversity plan to meet the needs of their ever changing ethnically diverse, disadvantaged, and multilingual population, (2) what process did the district use in meeting the needs of a diverse population, and (3) can a business model be transformed and used in an educational system to meet the needs of the internal forces as well as those external forces. The Cox’s Model for Cultural Change was used to analyze the district diversity plan. The district diversity plan had six categories while the Cox Model has five. According to the Cox Model, to have an effective organizational change the change effort should include all the five themes and sub- themes. The change effort should be continually assessed and refined over time in a process of continuous loop learning. A qualitative case study methodology was used to collect and report the data for this study. The data collected in this study included intensive open-ended and follow-up interviews. The data sources were comprised of one board member, three administrators, four principals, and eight teachers. Of the sixteen participants, six members were of color and ten were European-American teachers, principals, and administrators. The constant comparative method was used to analyze the data. In analysis, the themes utilized for this study, comparing the leadership of the board member, administrators, principals, and teachers, congruence in leadership as described by Cox did not exist. Failure to utilize all the elements of the Cox Model made Mayflower ISD Diversity Plan weak. Lack of communication across the district hampered Mayflower ISD to embrace diversity. The finding from this study using the three categories Leadership, Education, and Follow-Up revealed that it was lacking in continuity, understanding, implementation, guidance, and trust. In conjunction with the review of literature and an analysis of the data, the findings from this study reveal that the district will not meet its’ overall objective of embracing diversity district wide.

The reflectarray antenna has been considered as a suitable candidate to replace the traditional parabolic reflectors because of its high-gain and low-profile features. Beam scanning capability and multi-band operation are the current trends of the reflectarray design. It is desired to implement these functionalities with simple and effective techniques. Narrow bandwidth is the main issue which restricts the applications of the microstrip antennas. New microstrip slot antennas and polarization diversified planar antennas are introduced as the solutions to the issue of narrow bandwidth in this dissertation. A dual-band beam scanning reflectarray has been developed. It is the first offset-fed reflectarray that has been ever practically developed to emulate a cylindrical/parabolic type of reflector. Unlike other beam scanning reflectarrays which integrate phase tuning devices into the reflectarray elements and control the reflection phase, the beam scanning capability of this reflectarray is provided by its feed array. This method significantly reduces the complexity of the design of the beam scanning reflectarray. A new dual-band reflectarray configuration is also developed to eliminate the possible top layer blocking effects in the dual-layer reflectarray configuration. Perforated patches loaded with slots on the ground plane and rectangular patches loaded with slots on the patches are adopted as the low and high frequency bands, respectively. It is guaranteed that no physical contact between any two elements will occur. The bandwidth of the conventional microstrip antenna is small. A new wideband circularly polarized microstrip slot antenna is introduced in this dissertation. Very wide 3-dB axial ratio bandwidth is observed for the proposed antenna. The antennas are assembled in triangularly arranged array with sequential rotation feed technique. Polarization polarity is an alternative solution to the narrow bandwidth. A reconfigurable circularly polarized microstrip antenna is proposed. The antenna has both right-hand and left-hand circular polarizations which are controlled by two piezoelectric transducers. In addition, a dual-band dual-linearly-polarized planar array is designed based on the concepts of polarization diversity and multi-band operation. The research presented in this dissertation suggests useful techniques for reflectarrays and novel antenna designs. The results should have many applications for the modern wireless communication and radar systems.

SNP (single nucleotide polymorphisms, pronounce as snip) is one nucleotide position difference within human population. These differences can be detected in human genome and the difference occurs once about every 1000 base pairs. There are only two possible nucleotides in each SNP position. As a genetic marker, SNP data can be used to capture human disease traits because of its abundance and low diversity. In recent research results, it has been shown that there is a block-like structure in human genome, and only limited haplotype diversity can be observed. Consequently, we can use only a small fraction of SNPs to capture haplotype diversity in each block, and these SNPs are called tagSNPs. We propose a fixed-diversity approach to capture the diversity of the entire data. After partitioning the haplotype blocks, we will provide an objective way for evaluating the result. We obtain that the diversity of chromosome 21 SNPs locates at 0.5 by using our algorithm. The partition result shows the concurrence property of the haplotype data downloaded from NCBI web site. Finally, we develop an algorithm for tagSNP selection within each block, and obtain the compression ratio 0.78.

Large scale and planned introduction of maize (Zea mays) in southern Africa was accomplished during the last 100 years. Since then, smallholder farmers and breeders have been selecting varieties best adapted to their specific growing conditions. Six studies were conducted to generate information on the current levels of genetic diversity and agronomic performance of both farmer-developed and commercially-bred maize varieties in Zimbabwe, Zambia and Malawi to help in the identification of sources of new alleles for improving yield, especially under the main abiotic stresses that prevail in the region. In the first study, 267 maize landraces were collected from smallholder farmers in different agro-ecological zones of the three countries for conservation and further studies. Passport data and information on why smallholder farmers continue to grow landraces despite the advent of modern varieties were also collected along with the landraces. The second study revealed considerable variation for phenological, morphological and agronomic characters, and inter-relationships among the landraces and their commercial counterparts. A core sample representing most of the diversity in the whole collection of landraces was selected for further detailed analyses. The third study revealed high levels of molecular diversity between landraces originating from different growing environments and between landraces and commercially-bred varieties. The Simple Sequence Repeat (SSR) data also showed that the genetic diversity introduced from the original gene pool from the USA about 100 years ago is still found in both the descendant landraces and commercially-bred varieties. The fourth study showed that in general, commercially-bred varieties outyielded landraces under both abiotic stress and nonstress conditions with some notable exceptions. Landraces were more stable across environments than improved varieties. The most promising landraces for pre-breeding and further investigation were also identified. The clustering patterns formed based on agronomic data were different from SSR markers, but in general the genotype groupings were consistent across the two methods of measuring diversity. In the fifth study, the more recently-bred maize varieties in Zimbabwe showed consistent improvement over older cultivars in grain yield. The apparent yearly rate of yield increase due to genetic improvement was positive under optimum growing conditions, low soil nitrogen levels and drought stress. The sixth study revealed that in general, genetic diversity in Zimbabwean maize has neither significantly decreased nor increased over time, and that the temporal changes observed in this study were more qualitative than quantitative. The results from the six studies confirm the origin of maize in southern Africa and reveals that considerable genetic variation exists in the region which could be used to broaden the sources of diversity for maize improvement under the current agro-ecological conditions in southern Africa.

The urban Influence on diversity of avifauna in the Edwards Plateau ecoregion and surrounding area was studied using spatial analysis. Indices and metrics of urban influence, ownership property sizes, landscape structure, and avian diversity were calculated for 31 North American Breeding Bird Survey (BBS) transects, 12 located within the Edwards Plateau ecoregion and 18 in contiguous ecoregions. Spatial correlations were calculated between each pair of these indices. The spatial analysis identified an emergent property at the landscape level: A “threshold of habitat fragmentation” at an ownership property size of 500 acres, which is reached when urban influence increases to an intermediate level. Highly significant spatial correlations among variables showed that property sizes lower than 500 acres produce habitat fragmentation represented by a decrease in mean patch size (MN) and proximity among habitat patches (Index PROX). Consequently, avian α-diversity (richness) decreases because both MN and Index PROX are landscape metrics related to availability of suitable habitat for avian populations. The spatial analysis also made possible the prioritization of ecological subregions of the Edwards Plateau for conservation or restoration with respect to the threshold of habitat fragmentation and avian α and β-diversity. Balcones Canyon Lands showed a high percentage of land covered by farms smaller than 500 acres (64%), an ownership property average size above the threshold of fragmentation (1440 acres) and the highest avian α- diversity; so, management policies should focus on habitat conservation. In contrast, Lampasas Cut Plains showed the highest percentage of land covered by farms smaller than 500 acres (71%), and ownership property average size was very close to the threshold of fragmentation (625 acres); there, urban bird species are dominant and avian α-diversity is low because of the loss of native bird species. Management in this ecoregion should focus on habitat restoration. Finally, the Live Oak-Mesquite Savannah subregion showed the highest average ownership property size (7305 acres), and the highest values of patch richness and β-diversity. Management in this ecoregion should focus on conservation of land mosaic diversity to assure native avian species turnover.

The ecological forces that drive community structure of deep-sea benthic communities are poorly understood, yet such communities rival in biological complexity those of coral reefs or rainforests. Using components of the recently concluded DGoMB project, local and regional-scale structure of benthic macrofaunal communities were examined at thirty two locations throughout the continental slope of the northern Gulf of Mexico. Controlling factors associated with sediment disturbance, food supply, and faunal competition between functional ecological groups were evaluated for correlative and relational patterns. A higher order taxonomic sufficiency approach was used to calculate both alpha and beta diversity. The results of this study indicate that macrofaunal communities are very patchy, having wide variations in abundance at within-site, adjacent-site, and across-basin scales, yet all sample areas possess a large richness of higher taxa. Declining abundance was noted with increasing water depth and reduced particulate organic carbon levels. Upper-slope submarine canyons possess some of the highest abundances. Less mobile macrofauna, such as poriferans, bivalves, and scaphopods, dominate slope communities above the 500 meter contour. Sediments exhibiting intense megafaunal bioturbation inhibit abundances of sedentary macrofaunal taxa, but such mixing is positively associated with increased abundances of polychaetes and ambulatory crustaceans, including peracarids, harpacticoids, and ostracods. Prominent sediment mixing was noted at most sites, including portions of the Sigsbee Abyssal Plain. The western Gulf of Mexico was less biologically active than the eastern Gulf of Mexico, which possesses two extensive submarine canyons that appear to act as regional nutrient traps. I conclude that the physiographic complexity of the northern Gulf of Mexico continental slope influences macrofaunal community structure. Biological disturbance, in the form of sediment mixing, is widespread throughout most slope depths, and the benthic environment is food-limited. It appears that disequilibrium-type ecological processes predominate in this area, supporting similar findings by previous studies in other regions of the ocean, usually at far smaller scales and none representative at the basin-level. Use of higher order taxonomy in lieu of genus or species-level faunal identifications for diversity measurements was inadequate for detecting spatial patterns or environmental responses.

none

Thesis--PlanB (M.S.)--University of Wisconsin--Stout, 2001. / Field problem. Includes bibliographical references.

Thesis (Ph. D.)--University of Queensland, 2002. / Includes bibliographical references.