Infections frequently occur on both networks of devices and networks of people, and can model not only viruses, but also information, rumors, and product use. However, in many circumstances, the infection process itself is hidden, and only the effects, e.g. sickness or knowledge, can be observed. In addition, this information is likely incomplete, missing many sick nodes, as well as inaccurate, with false positives. To use this data effectively, it is often essential to identify the infection process causing the sickness, or even whether the cause is an infection. For our purposes, we consider the susceptible-infected (SI) infection model. We seek to distinguish between infections and random sickness, as well as between different infection (or infection-like) processes in a limited information setting. We formulate this as a hypothesis testing problem, where (typically) in the null, the sickness affects nodes at random, and in the alternative, the infection is spread through the network. Similarly, we consider the case where the sickness may be caused by one of two infection (or infection-like) processes, and we wish to find which is the causative process. We do this is a setting with very limited information, given only a single snapshot of the infection. Only a small portion of the infected population reports the sickness. In addition, there are several other limitations we consider. There may be false positives, obfuscating the infection. Similarly, there may be a random sickness and epidemic process occurring simultaneously. Knowledge of the graph topology may be incomplete, with unknown edges over which the infection may spread. The graph may also be weighted, affecting the way the infection spreads over the graph. In all these cases, we develop algorithms to identify the causative process of the infection utilizing the fact that infected nodes will be clustered. We demonstrate that under reasonable conditions, these algorithms detect an infection with asymptotically zero error probability as the graph size increases. / text

published_or_final_version / abstract / Statistics and Actuarial Science / Master / Master of Philosophy

No description available.

No description available.

Grassland bird species have declined dramatically since 1966. This decline can be linked to changes in land use practices, such as grazing. I examined the effects of cattle grazing on the abundance of birds by testing the predictions of the More Individuals Hypothesis (MIH). The study was conducted in Grasslands National Park of Canada (GNPC) in Saskatchewan. Point counts were used to sample richness and relative abundance of birds. I sampled two groups of invertebrates: grasshoppers and carabid beetles. In addition, vegetation measurements were taken to assess the intensity of grazing. I found that (1) grasshopper abundance, richness and diversity were higher in grazed pastures; (2) carabids showed mixed responses to grazing; (3) bird abundance was correlated with carabid abundance, thus supporting the assumptions of the MIH. Overall, my results indicate that grazing can be beneficial for both birds and their invertebrate prey in southern Saskatchewan mixed-grass prairies.

In statistical inference one of the most important properties that an estimator of an unknown parameter can possess is the property of sufficiency. The use of sufficient statistics has been prescribed in the past and their use today plays an ever increasing role in modern statistical inference. Because of the utmost importance of sufficient statistics, it is the goal of this thesis to study the various aspects of sufficiency in connection with the estimation of parameters.This study utilizes a blend of a geometric and an analytic approach to sufficiency. This is done to show the power of each and to demonstrate how the two approaches complement each other and combine to give better insight into the concept of sufficiency.In this thesis the author also establishes some new and important results on sufficiency and minimal sufficiency. These results, taken as a whole, constitute a unified presentation of some of the most important aspects of sufficiency and non-sufficiency.

Grassland bird species have declined dramatically since 1966. This decline can be linked to changes in land use practices, such as grazing. I examined the effects of cattle grazing on the abundance of birds by testing the predictions of the More Individuals Hypothesis (MIH). The study was conducted in Grasslands National Park of Canada (GNPC) in Saskatchewan. Point counts were used to sample richness and relative abundance of birds. I sampled two groups of invertebrates: grasshoppers and carabid beetles. In addition, vegetation measurements were taken to assess the intensity of grazing. I found that (1) grasshopper abundance, richness and diversity were higher in grazed pastures; (2) carabids showed mixed responses to grazing; (3) bird abundance was correlated with carabid abundance, thus supporting the assumptions of the MIH. Overall, my results indicate that grazing can be beneficial for both birds and their invertebrate prey in southern Saskatchewan mixed-grass prairies.

Cross-cultural exchange between individuals, both through face-to-face encounter and, more recently, electronic media, has been widely promoted as a means of educating against prejudice. Through the integration of structured field study, practical experience, educational and social theory, and contemporary philosophy, this thesis challenges the very foundations upon which such initiatives are built, and develops an alternative basis from which to approach cross-cultural exchange. An exploration of social categorisation posits the foundations of prejudice in essentialist conceptualisations, whether under the umbrella of universalism or relativism. The historical propensity for antiracist and multicultural education to reify group difference and reinforce such essentialist conceptualisations of identity thus presents an interesting conundrum. The fact that similar tendencies are noted in contemporary practice, educational resources, and official guidance, gives this more than an academic interest. Despite the apparent advantages of abandoning essentialist categorisations, studies of communication and identity formation reveal contradictory evidence - the need to locate others socially and to predict their behaviour accordingly, both in face-to-face and electronic communication. Thus, the challenge for educators is to develop innovative pedagogical approaches that translate contemporary, non-essentialist, understandings of group categorisation into workable practices to overcome inequality. While, within such a programme, cross-cultural exchange might be seen to have a valuable role to play alongside structural reform, it is clear that alone it cannot provide a panacea for prejudice. In the light of the above, the thesis addresses guidance for good practice in crosscultural exchange, and the related complexity of programme evaluation, alongside the training of future facilitators of such projects. A particular emphasis is placed upon the use of participatory arts and the unique tools this medium can bring to inclusive, cross-cultural collaboration. The thesis demonstrates, that cross-cultural exchange has the potential to provide valuable and significant learning experiences, some of which have previously been given little recognition.

The main purpose of this research project was to test the social risk hypothesis of depression proposed by Allen and Badcock (2003). The social risk hypothesis suggests that mild to moderate depression has evolved to promote belonging in small communities by making members sensitive to signs of rejection and motivated to restore their social value. Using self-report data from 397 working adult participants, structural equation modeling (SEM) was used to examine the relationships between secure attachment, social comparison, defeat, self-esteem, depression, submissive behaviours, social risk taking, and interpersonal sensitivity. Two further studies were also performed. The first tested whether the social risk hypothesis could explain anxiety as well as depression. The second tested gender invariance within the models of depression and anxiety. The dissertation is organized into three papers, preceded by a general introduction and followed by a general conclusion. The first paper is focused on the general test of the social risk hypothesis, the second on the comparison test of anxiety, and the third on the role of gender in the models generated. In the first paper, the SEM analysis indicated a good fit between the data and Allen and Badcock's (2003) algorithmic model, providing empirical support for an evolved adaptive mechanism functioning in mild to moderate depression. Paper 2 reports a test of Allen and Badcock's (2003) claim that the social risk hypothesis is exclusive to depression. In general, the anxiety model provided a fairly good fit to the social risk hypothesis; however, anxiety did not mediate the relationship between secure attachment and the two outcome variables (i.e., interpersonal sensitivity and submissive behaviours) suggesting that depression and anxiety have not evolved to function in exactly the same way. The goal of Paper 3 was to examine how the variables within the social risk hypothesis might operate differently for men and women given that past research strongly indicates that gender may have differential effects on the depressive (or anxious) mechanism. Two differences were found in the gender invariance analysis: i) the relationship between secure attachment and depression was much stronger for men; ii) men and women differed on the relationship between social comparison and anxiety. The papers discuss the findings from the perspective of evolutionary theory. / Counselling Psychology

Thesis (M. Env. Sc.)--University of Adelaide, Mawson Graduate Centre for Environmental Studies, 1993. / Includes bibliographical references.