This research investigates human shrunken head specimens from a forensic art and anthropological perspective. ‘Ceremonial tsantsa’ refers to shrunken heads mummified as war trophies within the ancient traditions and rituals of the Shuar, Achuar, Awajún, Wampís and Candoshi (SAAWC), within the primitive conditions of the Amazon rainforest. ‘Commercial shrunken heads’ are comparatively modern, with the first known accounts commencing from 1872. Low earning individuals, outwith the SAAWC, who had access to corpses and appropriate medical or taxidermy provisions, produced them for trade. These heads were made in abundance and do not present the same historical value. A total of 65 shrunken heads were examined (20 from the Smithsonian Institute, USA, 44 from the Science Museum, UK, and one from the Elgin Museum, UK). The aim was to establish whether any aspect of facial appearance could be determined from these shrunken heads that might enable identification of the deceased. The objectives of this research include: 1) Create a catalogue of heads. 2) Distinguish ceremonial from commercial heads and ascertain other potential qualities that may differentiate them. 3) Examine the quality of preserved tissues and facial structures using: manual inspection, microscopic hair analysis, visible spectrum light photography, ultraviolet fluorescence (UVF), infrared reflectography (IRR), morphometrics and MDCT scans. 4) Generate average shrinkage patterns in relation to the preservation of facial features to produce an example demonstrating the changes observed from living facial appearance to a shrunken head. 5) Produce multiple facial depictions from the shrunken heads to illustrate the likely living appearance of the victim and/or to highlight key features that could be used for identification. A total of 6 ceremonial tsantsa and 36 commercial heads were identified. Greater confidence is prescribed to the assigned commercial heads because their morphological appearance distinctly opposes the highly standardised condition of ceremonial tsantsa. Many indicated that the processor had access to modern resources such as gloves and fine suturing equipment, which was not typically available to the SAAWC. Due to traders sometimes closely replicating ceremonial tsantsa when shrinking and decorating heads for trade, this limits the certainty when prescribing heads to this category. Minor deviations in ceremonial design resulted in 23 heads being defined ambiguous. New ceremonial/commercial differentiating characteristics, identified directly from this research, include: • Ceremonial tsantsa present consistent epidermal degeneration; use of over and over sutures to repair skull removal incisions; distinct pointing of the nostril shape from manual manipulation; a smooth, even neck base, from slicing away the supportive neck ring post-desiccation; frequent application of strong compressions at the temples and lateral margins of the posterior head; stresses from the heavy remodelling of tissues demonstrated a tendency for symptoms recognised in the leather industry as ‘double hiding’ and ‘drawn grain’. • Commercial heads were identified to retain the epidermis in 64% of cases; 72% favoured ‘baseball’ stitches for suturing incisions; nostril shape was less distorted with 43% presenting only minor pointing; 64% did not refine the neck base post-desiccation, thus leaving a ragged, sometimes concertinaed edge; 53% had facial down trimmed or shaved, not singed. According to microscopic hair analysis, the majority of heads were adult (95%) and Mongoloid (75%). Juvenile heads and other ancestry groups were identified only in the commercial category. Sex estimation was limited to superficial male attributes. Evident facial hair growth identified 25% of heads as most likely male, predominantly in the commercial heads. Tertiary facial hair follicles, identified using UVF and IRR, gauged 15% as possibly male, predominantly in ceremonial tsantsa due to an improved visibility of the dermis texture. The condition and pathologies presenting in hair may indicate the environment in which heads existed, or the type of handling incurred. Evident sun bleaching, fungal infections or fractures can however occur either ante-mortem or post-mortem. The most reliable features identified in ceremonial and commercial specimens, to aid facial recognition, are skin pathologies, facial creases, hairlines and earlobe form. Hair pathologies that can only accumulate on living scalps may also be retained (e.g. nits and hair casts). The cartilages of the ear and nose, also retain some relevant morphological information, but are prone to alteration. Commercial heads are less distorted than ceremonial tsantsa and more suited to post-mortem depiction (PMD), often offering additional information, including eyebrow shape, vermillion border/lip shape, lip thickness (if mouth open), philtrum form, and possibly the palpebral slit angle. Unreliable features, as a result of skull removal, include loss of outer face shape and cheekbone structure, angle of projection for the ears and nose and degree of evident prognathism. Natural skin complexion is lost due to the depletion of melanin, epidermal degeneration and any applied stains. Hair appears longer (unless cut) and more profuse following shrinkage.

Master of Science / Department of Statistics / Haiyan Wang / Cancer type classification with high throughput molecular data has received much attention. Many methods have been published in this area. One of them is called PAM (nearest centroid shrunken algorithm), which is simple and efficient. It can give very good prediction accuracy. A problem with PAM is that this method selects too many genes, some of which may have no influence on cancer type. A reason for this phenomenon is that PAM assumes that all genes have identical distribution and give a common threshold parameter for genes selection. This may not hold in reality since expressions from different genes could have very different distributions due to complicated biological process. We propose a new method aimed to improve the ability of PAM to select informative genes. Keeping informative genes while reducing false positive variables can lead to more accurate classification result and help to pinpoint target genes for further studies. To achieve this goal, we introduce variable specific test based on Edgeworth expansion to select informative genes. We apply this test on each gene and select some genes based on the result of the test so that a large number of genes will be excluded. Afterward, soft thresholding with cross-validation can be further applied to decide a common threshold value. Simulation and real application show that our method can reduce the irrelevant information and select the informative genes more precisely. The simulation results give us more insight about where the newly proposed procedure could improve the accuracy, especially when the data set is skewed or unbalanced. The method can be applied to broad molecular data, including, for example, lipidomic data from mass spectrum, copy number data from genomics, eQLT analysis with GWAS data, etc. We expect the proposed method will help life scientists to accelerate discoveries with highthroughput data.

Doctor of Philosophy / Department of Statistics / Haiyan Wang / This dissertation considers different methods to improve the performance of the Prediction Analysis of Microarrays (PAM). PAM is a popular algorithm for high-dimensional classification. However, it has a drawback of retaining too many features even after multiple runs of the algorithm to perform further feature selection. The average number of selected features is 2611 from the application of PAM to 10 multi-class microarray human cancer datasets. Such a large number of features make it difficult to perform follow up study. This drawback is the result of the soft thresholding method used in the PAM algorithm and the thresholding parameter estimate of PAM. In this dissertation, we extend the PAM algorithm with two other thresholding methods (hard and order thresholding) and a deep search algorithm to achieve better thresholding parameter estimate. In addition to the new proposed algorithms, we derived an approximation for the probability of misclassification for the hard thresholded algorithm under the binary case. Beyond the aforementioned work, this dissertation considers the heteroscedastic case in which the variances for each feature are different for different classes. In the PAM algorithm the variance of the values for each predictor was assumed to be constant across different classes. We found that this homogeneity assumption is invalid for many features in most data sets, which motivates us to develop the new heteroscedastic version algorithms. The different thresholding methods were considered in these algorithms. All new algorithms proposed in this dissertation are extensively tested and compared based on real data or Monte Carlo simulation studies. The new proposed algorithms, in general, not only achieved better cancer status prediction accuracy, but also resulted in more parsimonious models with significantly smaller number of genes.

This thesis examines the rhetorical and visual manifestations of race as they figured in the months prior to and within the 2006 New Orleans mayoral election discourses, and examines how the Nagin campaign tapped into a strategy that capitalized upon pre-existing racial tensions exacerbated by Katrina in order to win re-election. Much of the research for this thesis emerged from the making of a documentary film that examines the intersection between race and politics within this same election, and draws upon primary source video interviews conducted between February - May, 2006, and secondary source media and communications materials to posit that race rendered all political response to Katrina impotent, and that the reductive discourse of a racialized campaign was founded upon traditional, outmoded, and predictable interpretations of racial differences facilitated by socioeconomic hierarchies that both provided a structure for and allowed the psychological framework for such a strategy to work.