Thesis (Ph.D.)--Case Western Reserve University, 2009 / Department of Biomedical Engineering Abstract Title from OhioLINK abstract screen (viewed on 10 April 2009) Available online via the OhioLINK ETD Center

Thesis (Ph. D.)--Drexel University, 2005. / Includes abstract and vita. Includes bibliographical references (leaves 72-80).

We start by showing that in an active learning setting, the Perceptron algorithm needs $\Omega(\frac{1}{\epsilon^2})$ labels to learn linear separators within generalization error $\epsilon$. We then present a simple selective sampling algorithm for this problem, which combines a modification of the perceptron update with an adaptive filtering rule for deciding which points to query. For data distributed uniformly over the unit sphere, we show that our algorithm reaches generalization error $\epsilon$ after asking for just $\tilde{O}(d \log \frac{1}{\epsilon})$ labels. This exponential improvement over the usual sample complexity of supervised learning has previously been demonstrated only for the computationally more complex query-by-committee algorithm.

Διάφορες μορφές κωδικοποίησης Huffman έχουν προταθεί για τη συμπίεση των δεδομένων δοκιμής που χρησιμοποιούνται για τον έλεγχο της ορθής λειτουργίας ολοκληρωμένων συστημάτων μετά την κατασκευή τους. Μεταξύ αυτών η βέλτιστη επιλεκτική κωδικοποίηση Huffman παρουσιάζει διάφορα πλεονεκτήματα. Η υιοθέτηση διαφορετικής τεχνικής ελέγχου της ορθής λειτουργίας του ολοκληρωμένου συστήματος μετά την κατασκευή του και στο πεδίο της εφαρμογής αυξάνει το κόστος του συστήματος. Για το λόγο αυτό στην εργασία αυτή διερευνούμε τη δυνατότητα χρησιμοποίησης της βέλτιστης επιλεκτικής κωδικοποίησης τόσο μετά την κατασκευή του ολοκληρωμένου συστήματος όσο και στο πεδίο της εφαρμογής. Τα συστήματα που μας ενδιαφέρουν είναι τα ενσωματωμένα συστήματα πραγματικού χρόνου. Για το λόγο αυτό μελετάμε καταρχήν τα χαρακτηριστικά που πρέπει να έχει η τεχνική ελέγχου ορθής λειτουργίας στο πεδίο της εφαρμογής ανάλογα με τις απαιτήσεις ενός ενσωματωμένου συστήματος. Σε κάθε περίπτωση μελετάμε το κόστος υλοποίησης της τεχνικής ελέγχου τόσο σε κυκλώματα, σχεδιάζοντάς τα σε Verilog, όσο και στο βαθμό συμπίεσης των δεδομένων δοκιμής, κάνοντας εξομοιώσεις σε C. / One-time factory testing of VLSI component after fabrication is insufficient in the deep submicron area. The products must be tested periodically in the field of application. Due to the complexity of the Systems on a Chip (SOCs), huge amounts of test data are required. However in many embedded systems the capacity of the available memory is a limited resource. In Automatic Test Equipment (ATE) based factory testing in order to reduce the memory requirements of the ATE and the time required to transfer the test data from ATE into the chip (and hence the test application time) various test set compression techniques have been proposed. In this paper we investigate the required enhancements so that an Optimal Selective Huffman Coding based test-set compression technique can be used for periodic testing in the field. The requirements of various types of periodic testing are examined depending on the criticality of the application running in embedded systems.

Many complex decisions require integrating and assessing multiple streams of dynamic information whilst determining how to act. This dynamic information often contains rhythmic structures which our sensory systems can adapt to and use to anticipate future events. Despite the close relationship between rhythmic temporal expectations and complex decision making being self evident, no experiments explicitly attempt to understand this interdependence. If the theories that have emerged from both domains are to generalise to complex interactive behaviour, the effects of dynamic context on complex decisions must be considered. I argue that timing research must move beyond simple decisions and develop a new experimental framework for addressing the problem. This includes increasing the complexity of experimental tasks, testing the effects of timing on perceptual averaging and subjective value decisions, incorporating timing as an inherent dimension of targets, testing degrees of aperiodicity and exploring the effects that prior knowledge about the temporal structure of a stimulus has on choice. Seven behavioural experiments are reported that implement the new experimental framework. Five use a complex auditory-spatial averaging task to examine effects of periodicity, expectation, prior knowledge and related parameters such as IOI variance. One tests the effects of rhythmic variability and stimulus duration on auditory detection to determine specificity to complex decision making, and one investigates the effects of timing on audio-visual subjective value decisions. The results show that existing theories of temporal expectation do not necessarily generalise to complex decision making. Periodicity reduces the amount of information that is needed to form complex decisions. However, the effects of periodicity (or degree of aperiodicity) on choice are dependent on a number of factors associated with prior knowledge, stimulus rate, variance, decision type and task complexity. Using these findings I develop an explanatory framework called "dynamic inhibition and boosting" that better accounts for behavioural data in the literature compared with existing theories. This explanation is supported by the novel proposal that temporal expectations influence confidence and perceived risk.

xi, 103 p. : ill. A print copy of this thesis is available through the UO Libraries. Search the library catalog for the location and call number. / Although outcrossing is the most widespread mating system among animals and plants, the reason for this prevalence is not fully understood. Evolutionary theory has classified the potential selective pressures driving the evolution and maintenance of outcrossing into two broad categories: deleterious mutations and changing ecological conditions. Despite the inherent advantages of self-fertilization, exposure to either or both of these selective pressures is predicted to favor outcrossing over self-fertilization. I tested these predictions using experimental evolution in populations of Caenorhabditis elegans with genetically modified rates of outcrossing and selfing. I found that outcrossing reduces the fixation of deleterious mutations under mutation influx and that outcrossing expedites adaptation to a bacterial pathogen. Further, I identified facultative outcrossing, a novel life history characteristic, in specific C. elegans strains that predominantly reproduce by selfing but engage in outcrossing when stressed. The shift from a primarily selfing mating system to a predominantly outcrossing system is similar to the environmentally induced facultative sex observed in asexual species, which is thought to enable more rapid adaptation. Facultative outcrossing, although not previously documented, may play a major role in the life histories of many highly selfing species. Finally, most mutations are deleterious and therefore elevated mutation rates are generally thought to produce progressively larger reductions in fitness. Using the chemical mutagen ethylmethanesulfonate, I found the surprising result that populations exposed to a mutation rate at least fifty times greater than natural rates exhibited significantly greater fitness than populations exposed to substantially lower mutation rates. This unexpected fitness optimum may be the result of a volatile balance between the influx of deleterious mutations and compensatory mutations. This work confirms the predictions of several long-standing evolutionary theories by identifying both deleterious mutations and changing ecological conditions as selective pressures capable of driving the evolution and maintenance of outcrossing. These selective pressures, which are ubiquitous in nature, may explain the prevalence of outcrossing relative to selling. This dissertation includes previously published and co-authored materials. / Committee in charge: Barbara Roy, Chairperson, Biology; Patrick Phillips, Advisor, Biology; Karen Guillemin, Member, Biology; William Bradshaw, Member, Biology; Douglas Kennett, Outside Member, Anthropology

Selective attention refers to the ability to enhance the processing of relevant stimuli, while suppressing the processing of irrelevant distractors. The neural mechanisms of selective attention are vulnerable in children from lower socioeconomic status families, yet these neural mechanisms can also be enhanced with evidence-based, targeted training. The series of studies presented in this dissertation investigated the individual differences in development and neuroplasticity of selective auditory attention in association with nonverbal cognitive abilities, in relation to genetic influences, and in the context of gene x intervention interactions. To this end, a multi-method approach was adopted, combining several methodologies such as event-related potentials (ERPs), behavioral measures, molecular genetics, and a randomized, controlled intervention design. In the first study, individual differences in neural mechanisms of selective auditory attention were studied, in association with nonverbal cognitive abilities. More robust ERP selective attention effects were associated with superior nonverbal IQ performance. These results indicated a noteworthy relationship between neural mechanisms of selective attention and nonverbal IQ performance in lower socioeconomic status (SES) preschoolers. In the second study, the relationship between 5-HTTLPR polymorphism and neural mechanisms of selective auditory attention was assessed. ERPs of selective attention effect were larger in children who carried at least one short allele of 5-HTTLPR, in comparison to long-homozygotes. These results associated being homozygous for the long allele with weaker neural mechanisms of selective attention in lower SES children. In the third study, these genetic influences were investigated in the context of an effective family-based training program previously shown to improve neural mechanisms of selective attention in lower SES preschoolers. The long-homozygote children, who initially displayed more attenuated ERPs of selective auditory attention than their short-carrier peers, showed robust ERPs of selective attention at posttest, but only if they were randomly assigned to the training program. These findings demonstrated that an effective family-based training could moderate the genetic influences of 5-HTTLPR on the neural mechanisms of selective attention. Taken together, the studies presented in this dissertation contribute to elucidating individual differences in development and neuroplasticity of selective auditory attention in lower SES preschoolers. This dissertation includes unpublished co-authored material.

Master of Science / Department of Anatomy and Physiology / A. Philine Wangemann / SLC26A4 and its murine ortholog Slc26a4 code for pendrin, an anion-exchanger that is expressed in the inner ear. Patients with mutations in SLC26A4 have syndromic or nonsyndromic hearing loss that is associated with a prenatal enlargement of the membranous labyrinth. The mouse model Slc26a4[superscript]Δ/Δ recapitulates the enlargement, develops an enlargement of the inner ear and fails to acquire hearing. The vestibular labyrinth secretes fluid, accounting for enlargement of the membranous labyrinth. The objective of the current study was to measure K⁺ concentrations in the utricular endolymph of Slc26a4[superscript]Δ/+ and Slc26a4[superscript]Δ/Δ mice as a first step toward a mechanistic understanding of fluid secretion during perinatal development. Doublebarreled K⁺-selective electrodes were used to measure K⁺ concentrations of the utricular endolymph in vitro. Potassium concentrations were ~10 mM in both genotypes at embryonic (E) day 16.5. The K⁺ concentrations started to rise at E17.5 in Slc26a4[superscript]Δ/+ mice. There was a 1-day delay in Slc26a4[superscript]Δ/Δ mice. This delay may be the consequence of a larger fluid volume. K⁺ concentrations rose to 150 mM and 132 mM in Slc26a4[superscript]Δ/+ and Slc26a4[superscript]Δ/Δ adult mice, respectively. Consistently, expression of KCNQ1 and the Na⁺/2Cl⁻/K⁺ cotransporter SLC12A2 was found in the utricle at E19.5 in Slc26a4[superscript]Δ/+ and Slc26a4[superscript]Δ/Δ mice. In conclusion, the data suggest that K⁺ secretion is not the major driving force of fluid secretion in the utricle of the developing mouse inner ear.

Made available in DSpace on 2014-10-09T12:38:53Z (GMT). No. of bitstreams: 0 / Made available in DSpace on 2014-10-09T14:01:57Z (GMT). No. of bitstreams: 1 02809.pdf: 6234365 bytes, checksum: 25107a337d71b51089353992495839d5 (MD5) / Dissertacao (Mestrado) / IPEN/D / Instituto de Pesquisas Energeticas e Nucleares - IPEN/CNEN-SP

Submitted by MIRELA PERONI GARCIA null (mirelapgarcia@hotmail.com) on 2017-01-25T23:38:22Z No. of bitstreams: 1 Dissertação Mirela_Peroni_Garcia.pdf: 1229443 bytes, checksum: 9a5120835fc31152c5ae462121dcb313 (MD5) / Approved for entry into archive by LUIZA DE MENEZES ROMANETTO (luizamenezes@reitoria.unesp.br) on 2017-01-27T15:36:52Z (GMT) No. of bitstreams: 1 garcia_mp_me_jabo.pdf: 1229443 bytes, checksum: 9a5120835fc31152c5ae462121dcb313 (MD5) / Made available in DSpace on 2017-01-27T15:36:52Z (GMT). No. of bitstreams: 1 garcia_mp_me_jabo.pdf: 1229443 bytes, checksum: 9a5120835fc31152c5ae462121dcb313 (MD5) Previous issue date: 2016-11-28 / Um novo sistema de plantio de cana-de-açúcar, utilizando mudas-pré-brotadas (MPB) em substituição aos fragmentos de colmos de cana fundamenta a hipótese deste trabalho de que poderá haver intoxicação nas mudas pelos herbicidas que são aplicados em pré emergência da cultura no sistema tradicional. Portanto, objetivou-se neste estudo verificar a seletividade de tratamentos de herbicidas aplicados em pré-plantio da MPB de cana-de-açúcar. Foi realizado um experimento a campo, no delineamento em blocos casualizados, utilizando sete tratamentos (T1: 720 g i.a.ha-1 oxifluorfen + 600 g i.a.ha-1 sulfentrazone; T2: 720 g i.a. ha-1oxifluorfen + 2500 g i.a. ha-1 atrazina; T3: 1100 g i.a. ha-1 tebutiuron + 600 g i.a. ha-1 sulfentrazone; T4: 1920g i.a. ha-1 s-metolaclor + 600g i.a. ha-1sulfentrazone; T5: 1008 g i.a. ha-1 clomazone + 2500 g i.a. ha-1 atrazina; T6: 1080 g i.a. ha-1 clomazone + 600 g i.a. ha-1 sulfentrazone; T7: testemunha capinada) em 4 repetições, na cultivar CTC14. Após a sulcação, foi realizada a pulverização dos herbicidas e, após 24 horas, o plantio das mudas. No início do desenvolvimento das mudas todos os tratamentos com herbicida causaram fitointoxicação, mas com o crescimento e desenvolvimento das mudas os sintomas foram diminuindo, não havendo diferença de altura, diâmetro de colmos, número de folhas, eficiência quântica do FII (Fv/Fm) e matéria seca das plantas entre os tratamentos como, também, na produtividade e características tecnológicas dos colmos não houve diferença significativa. Todos os herbicidas utilizados foram seletivos às mudas de cana-de-açúcar no sistema MPB quando aplicados em pré-plantio. / A new sugarcane planting system, using seedlings, pre-sprouted (MPB) to replace the cane stalks substantiates the hypothesis of this study that there might be intoxication the seedlings by herbicides that are sprayed in pre emergence of the crop in traditional system. Therefore, the aim of this work was to study the selectivity of herbicides treatments applied in pre-planting “MPB” sugarcane. An experiment in the field was conducted in a randomized block design, using seven treatments (T1: 720 g i.a. ha-1 oxifluorfen + 600 g i.a. ha-1 sulfentrazone; T2: 720 g i.a.ha-1 oxifluorfen + 2500 g i.a. ha-1 atrazina; T3: 1100 g i.a. ha-1 tebutiuron + 600 g i.a. ha-1 sulfentrazone; T4: 1920 g i.a. ha-1 s-metolaclor + 600 g i.a. ha-1 sulfentrazone; T5: 1008 g i.a. ha-1 clomazone + 2500 g i.a. ha-1 atrazina; T6: 1080 g i.a. ha-1 clomazone + 600 g i.a. ha-1 sulfentrazone; T7: weeded control) in 4 replications in a commercial variety, CTC14. After furrow herbicides were sprayed and 24 hours after, planting the seedlings. Early in the development of seedlings all herbicide treatments showed phytotoxicity, but with the growth and development of culture the symptoms have diminished, with no height difference, stalk diameter, number of leaves, quantum efficiency of FII (Fv / Fm) and dry matter, between the treatments and in the harvest, there was no significant difference between the average productivity and technological characteristics to the stalks. All herbicides treatments were selective when applied in pre-planting to “MPB” system.