No description available.

This thesis investigates the use of newly-developed survival analysis tools for credit scoring. Credit scoring techniques are currently used by financial institutions to estimate the probability of a customer defaulting on a loan by a predetermined time in the future. While a number of classification techniques are currently used, banks are now becoming more concerned with estimating the lifetime of the loan rather than just the probability of default. Difficulties arise when using standard statistical techniques due to the presence of censoring in the data. Survival analysis, originating from medical and engineering fields, is an area of statistics that typically deals with censored lifetime data. The theoretical developments in this thesis revolve around linear regression for censored data, in particular the Buckley-James method. The Buckley-James method is analogous to linear regression and gives estimates of the mean expected lifetime given a set of explanato ry variables. The first development is a measure of fit for censored regression, similar to the classical r-squared of linear regression. Next, the variable-reduction technique of stepwise selection is extended to the Buckley-James method. For the last development, the Buckley-James algorithm is altered to incorporate non-linear regression methods such as neural networks and Multivariate Adaptive Regression Splines (MARS). MARS shows promise in terms of predictive power and interpretability in both simulation and empirical studies. The practical section of the thesis involves using the new techniques to predict the time to default and time to repayment of unsecured personal loans from a database obtained from a major Australian bank. The analyses are unique, being the first published work on applying Buckley-James and related methods to a large-scale financial database.

Salmonella outbreaks associated with orange juices have been reported in the past. Though there have been studies on the internalization of Salmonella into oranges there is inadequate information on the ability of this pathogen to survive on orange surfaces, become internalized, and survive the low pH internal conditions. The objective of this work was to study the survival of Salmonella gaminara and Salmonella agona on oranges obtained from the field and retail outlets and investigate their attachment and internalization potential. These studies showed that oranges obtained from both the field and retail outlets harbored relatively high concentrations of aerobic heterotrophic bacterial populations. There were significant differences in the survival of Salmonella agona and Salmonella gaminara at 4??C, room temperature (25??C) and 37??C. Survival was highest at 37??C and lowest at 4??C for both Salmonella gaminara and Salmonella agona. Salmonella agona and Salmonella gaminara showed significant differences in recovery when the cells were treated with pH 4.0, 7.0 and 9.5 buffers. The internalization studies suggest that a negative temperature differential favors the internalization of Salmonella cells into the fruit. Significant differences in the internalization of Salmonella into field and market oranges were observed with more internalization in the field oranges as compared to the market oranges. These results suggest that to prevent Salmonella contamination of orange juices adequate pre-harvest protection against pathogen contamination and post-harvest cleaning and disinfection strategies need to be employed.

Because of the failure so far to find effective treatment for patients with advanced stages of melanoma, increasing efforts have been made to find prognostic factors identifying patients in the risk zone for development of metastasis. This thesis investigates the prognostic powers of a few selected serological and immunohistochemical biomarkers. In the first and second study, patients operated on for localized malignant melanoma were investigated regarding the prognostic impact of angiogenic serological markers and circulating levels of S100. We concluded that the S100 assays, especially S100BB, are potential biomarkers in patients with malignant melanoma, correlated to both survival and disease free survival. However, no such conclusion could be drawn from the first study, where we found no correlation to survival and investigated angiogenic markers. In the third and fourth study four new potential immunohistochemical biomarkers where investigated in collaboration with the Swedish Human Protein Atlas Program, and those where TRP-1, galectin-1, DLG5 and syntaxin-7. We found that TRP-1 correlated inversely with tumor stage and galectin-1 correlated to Ki-67. DLG5 showed a significant inverse correlation to Ki67 and the expression of STX7 was inversely correlated to tumor stage, suggesting that decreased expression is associated with more aggressive tumors. None of the investigated markers in study III and IV correlated with disease free survival or overall survival. In the fifth and last study, we examined the expression of SOX10, a transcription factor, in different melanocytic lesions. Also, a proliferation assay was carried out in a human melanoma cell line. The results reveal the presence of SOX10 in different melanocytic lesions, with a weak inverse correlation to survival and a significant inverse correlation to T-stage. A significant decrease in proliferation rate for SOX10 silenced cells was found and our data also suggests an increased migratory response in SOX10 silenced cells.

The use of probiotics and prebiotics has become a popular trend in the food industry. The main goal of this study was to produce a synbiot by encapsulating a probiotic and a prebiotic within a matrix that would provide sufficient protection to the probiotic against simulated gastric juice (SGJ). The ability of the probiotic, Bifidbacterium adolescentis, to grow on short chain fructooligosaccharides (FOS; DP 2-8, P95), inulin (DP 2-60, ST), and FOS/inulin mixture (DP 2-60, Syn), as well as glucose and a glucose-free maltooligosaccharide (MOS), were evaluated. Bifidobacterium adolescentis had a significantly higher specific growth rate on P95 (0.47 h-1), than glucose (0.40 h-1). Examination of the growth medium containing P95 and MOS by high performance anion exchange with pulsed amperometric detection (HPAE-PAD) revealed that B. adolescentis utilised the oligosaccharides to the same extent as the monosaccharides.<p> Bifidobacterium adolescentis was successfully encapsulated with and without P95 using extrusion and emulsion methods, at cell concentrations of 8-9 log colony forming units (CFU) mL-1. Capsules formed by the extrusion method with 1.0% alginate (AL), 4.0% pea protein isolate (PPI) + 0.5% AL, and 4.0% whey protein isolate (WPI) + 0.5% AL ranged in geometric mean diameter from 2.0 to 2.2 mm. Capsules formed by emulsion with 4.0% WPI + 0.5% AL had geometric mean diameter of 53 ìm. Extrusionbased encapsulated probiotics in either PPI + AL or WPI + AL showed improved survival in SGJ at pH 2.0 for 2.0 h with log CFU mL-1 reductions of 3.6 and 1.1, respectively. Free cells, AL extrusion-based and WPI + AL emulsion-based encapsulated probiotics showed no survival after 30 min in SGJ at pH 2.0. The addition of 1.0% (w/w) P95 to the PPI + AL capsules improved probiotic survival such that 1.0 log CFU mL-1 reduction was observed. The amount of P95 encapsulated ranged from 4.0 to 4.4 mg per gram of capsules.<p> The external surface of the PPI + AL capsules as examined by cold stage scanning electron microscopy (cryo-SEM) and atomic force microscopy (AFM) was smooth with the presence of pores ranging in diameter from 0.25 to 1.00 ìm. The addition of P95 to the capsules had no significant effect on surface roughness as measured by AFM, but significantly increased the external capsule thickness. The internal structure of the PPI + AL capsules examined by cryo-SEM revealed a porous honeycomb-like structure, with inner pore diameters ranging between 13.0 and 21.9 ìm. Probiotic cells were found to be randomly dispersed on the surface and in the interior of the honeycomb pores. In contrast, the prebiotic was found to be distributed throughout the capsule as observed by confocal laser scanning microscopy (CLSM), indicating that it would be readily available to the probiotic as a carbon source

This descriptive study was done 1) to explore and describe the proportional distribution of breast, cervical and colorectal cancers by stage (a proxy measure of availability, access, and utilization of secondary prevention strategies) in northern Saskatchewan First Nations and non-First Nations in comparison to southern Saskatchewan First Nations and non-First Nations; 2) to assess the impact of stage and age on the survival patterns for these cancers in northerners and First Nations whose survival patterns have been shown by previous research to be equal or poorer in comparison to southerners. Univariate and multivariate survival analyses were carried out to ascertain the impact of the different proportions of stage for each study group on survival. Stage at time of diagnosis is a proxy assessment of secondary prevention services, which include formal screening programs.<p> Data for this study was obtained from the Saskatchewan Cancer Registry, which has been maintaining cancer data since 1932. Cancer stage at time of diagnosis information is complete in the registry for different years for each cancer site. Hence data for breast cancer was for the years 1970 to 1995; cervical cancer data for the years 1980 to 1995; colorectal cancer data for the years 1990 to 1995. <p> The proportion of cancer cases for each site by TNM stage and age were compared among the four study groups. First Nation and northern populations were found to have a larger proportion of diagnoses at a later stage in comparison to the southern non-First Nation group. <p> Using Cox's proportional hazards model, both stage and age at time of diagnosis were found to be significant predictors of survival for all study groups. Age and stage adjusted relative risks were calculated and found to be significant in comparison to the southern non-First Nation group for cancer of the breast (RR =1.81 P=0.013). For cervical cancer the relative risk of dying of cervical cancer for southern First Nations in comparison to southern non­-First Nations was found to be 1.38 but this was not statistically significant (p = 0.097). For colorectal cancer, the relative risk of dying of colorectal cancer was found to be better for northern First Nations in comparison to southern non-First Nations (RR = 0.59), however this was not statistically significant (p = 0.45).<p> This study showed that despite adjusting for stage and age at time of diagnosis, there were still some unexplained differences in the survival pattern of northern First Nations, northern non-First Nations and southern First Nations in comparison to southern non-First Nations. Hypotheses as to what these unexplained differences are have been offered. These include differences in socio-economic status as well as availability, accessibility, attitudes towards and knowledge of secondary prevention strategies. Further study into these unexplained differences should be carried out.

Pre-marital cohabitation has become a mainstream practice among couples in the United States, yet initial empirical evidence demonstrated significant correlations between cohabitation and subsequent marital instability. Later studies disputed a causal relation and have attempted to show a weakened connection over time, but have themselves suffered from the use of exclusively older and/or unreliable data. This paper uses figures from the National Longitudinal Survey of Youth (1997-08) and the National Survey of Family Growth (2006-08) to provide an updated analysis of cohabitation’s effect on marriage over the past decade. Using proportional hazard and competing risk models, this paper confirms a now minimal impact of cohabitation on marriage stability, with possible exceptions for serial cohabitation and cohabitation begun prior to engagement. Moreover, this study finds that cohabitation overwhelmingly remains a step toward, rather than serious substitute for, marriage.

This descriptive study was done 1) to explore and describe the proportional distribution of breast, cervical and colorectal cancers by stage (a proxy measure of availability, access, and utilization of secondary prevention strategies) in northern Saskatchewan First Nations and non-First Nations in comparison to southern Saskatchewan First Nations and non-First Nations; 2) to assess the impact of stage and age on the survival patterns for these cancers in northerners and First Nations whose survival patterns have been shown by previous research to be equal or poorer in comparison to southerners. Univariate and multivariate survival analyses were carried out to ascertain the impact of the different proportions of stage for each study group on survival. Stage at time of diagnosis is a proxy assessment of secondary prevention services, which include formal screening programs.<p> Data for this study was obtained from the Saskatchewan Cancer Registry, which has been maintaining cancer data since 1932. Cancer stage at time of diagnosis information is complete in the registry for different years for each cancer site. Hence data for breast cancer was for the years 1970 to 1995; cervical cancer data for the years 1980 to 1995; colorectal cancer data for the years 1990 to 1995. <p> The proportion of cancer cases for each site by TNM stage and age were compared among the four study groups. First Nation and northern populations were found to have a larger proportion of diagnoses at a later stage in comparison to the southern non-First Nation group. <p> Using Cox's proportional hazards model, both stage and age at time of diagnosis were found to be significant predictors of survival for all study groups. Age and stage adjusted relative risks were calculated and found to be significant in comparison to the southern non-First Nation group for cancer of the breast (RR =1.81 P=0.013). For cervical cancer the relative risk of dying of cervical cancer for southern First Nations in comparison to southern non­-First Nations was found to be 1.38 but this was not statistically significant (p = 0.097). For colorectal cancer, the relative risk of dying of colorectal cancer was found to be better for northern First Nations in comparison to southern non-First Nations (RR = 0.59), however this was not statistically significant (p = 0.45).<p> This study showed that despite adjusting for stage and age at time of diagnosis, there were still some unexplained differences in the survival pattern of northern First Nations, northern non-First Nations and southern First Nations in comparison to southern non-First Nations. Hypotheses as to what these unexplained differences are have been offered. These include differences in socio-economic status as well as availability, accessibility, attitudes towards and knowledge of secondary prevention strategies. Further study into these unexplained differences should be carried out.

The use of probiotics and prebiotics has become a popular trend in the food industry. The main goal of this study was to produce a synbiot by encapsulating a probiotic and a prebiotic within a matrix that would provide sufficient protection to the probiotic against simulated gastric juice (SGJ). The ability of the probiotic, Bifidbacterium adolescentis, to grow on short chain fructooligosaccharides (FOS; DP 2-8, P95), inulin (DP 2-60, ST), and FOS/inulin mixture (DP 2-60, Syn), as well as glucose and a glucose-free maltooligosaccharide (MOS), were evaluated. Bifidobacterium adolescentis had a significantly higher specific growth rate on P95 (0.47 h-1), than glucose (0.40 h-1). Examination of the growth medium containing P95 and MOS by high performance anion exchange with pulsed amperometric detection (HPAE-PAD) revealed that B. adolescentis utilised the oligosaccharides to the same extent as the monosaccharides.<p> Bifidobacterium adolescentis was successfully encapsulated with and without P95 using extrusion and emulsion methods, at cell concentrations of 8-9 log colony forming units (CFU) mL-1. Capsules formed by the extrusion method with 1.0% alginate (AL), 4.0% pea protein isolate (PPI) + 0.5% AL, and 4.0% whey protein isolate (WPI) + 0.5% AL ranged in geometric mean diameter from 2.0 to 2.2 mm. Capsules formed by emulsion with 4.0% WPI + 0.5% AL had geometric mean diameter of 53 ìm. Extrusionbased encapsulated probiotics in either PPI + AL or WPI + AL showed improved survival in SGJ at pH 2.0 for 2.0 h with log CFU mL-1 reductions of 3.6 and 1.1, respectively. Free cells, AL extrusion-based and WPI + AL emulsion-based encapsulated probiotics showed no survival after 30 min in SGJ at pH 2.0. The addition of 1.0% (w/w) P95 to the PPI + AL capsules improved probiotic survival such that 1.0 log CFU mL-1 reduction was observed. The amount of P95 encapsulated ranged from 4.0 to 4.4 mg per gram of capsules.<p> The external surface of the PPI + AL capsules as examined by cold stage scanning electron microscopy (cryo-SEM) and atomic force microscopy (AFM) was smooth with the presence of pores ranging in diameter from 0.25 to 1.00 ìm. The addition of P95 to the capsules had no significant effect on surface roughness as measured by AFM, but significantly increased the external capsule thickness. The internal structure of the PPI + AL capsules examined by cryo-SEM revealed a porous honeycomb-like structure, with inner pore diameters ranging between 13.0 and 21.9 ìm. Probiotic cells were found to be randomly dispersed on the surface and in the interior of the honeycomb pores. In contrast, the prebiotic was found to be distributed throughout the capsule as observed by confocal laser scanning microscopy (CLSM), indicating that it would be readily available to the probiotic as a carbon source

Lesser prairie chicken (Tympanuchus pallidicinctus) numbers have declined considerably in Texas since the early 1900s. Conversion of native prairie to cropland has been the major cause of the decline. I trapped and monitored 115 (66 males, 49 females) lesser prairie chickens in the Rolling Plains of the Texas Panhandle from 2001 through 2003. I used an information-theoretic approach to model selection as implemented in program MARK to evaluate factors contributing to variation in survival and differences in nest success. I found breeding season survival of both males and females was lower compared to non-breeding season survival. Annual survival was 0.52 (95% CI: 0.32? 0.71). Model selection indicated higher nest success (70%) in the sand sagebrush (Artemisia filifolia) vegetation type as compared to the shinnery oak (Quercus harvardii) type (40%). I also evaluated post-burn habitat alterations and plant succession (1 year and 2 years after burning) as potential lesser prairie chicken habitat. After spring rainfalls stimulated re-growth of herbaceous plants, male lesser prairie chickens moved to the site, feeding on new-emerging forbs throughout the summer. A female lesser prairie chicken with a brood used the burned site during the first summer after the burn. A year later, males established a lek on the burned site. Two female lesser prairie chickens with broods used the burned site during the second summer. Burned sites had more forbs than nonburned sites and probably had more insects available which are an important food source for chicks during their first 4?5 weeks of age.