Thesis (Ph.D.)--Indiana University, Dept. of Psychological and Brain Sciences, 2007. / Source: Dissertation Abstracts International, Volume: 68-07, Section: B, page: 4891. Adviser: Steven J. Sherman. Title from dissertation home page (viewed Apr. 15, 2008).

Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 2007. / Source: Dissertation Abstracts International, Volume: 68-11, Section: B, page: 7701. Adviser: Charles H. Hillman. Includes bibliographical references (leaves 103-116) Available on microfilm from Pro Quest Information and Learning.

Thesis (Ph.D.)--Indiana University, Depts. of Psychological and Brain Sciences and Cognitive Science, 2008. / Title from PDF t.p. (viewed on Jul 22, 2009). Source: Dissertation Abstracts International, Volume: 69-10, Section: B, page: 6448. Advisers: Robert L. Goldstone; Peter M. Todd.

Thesis (M.S.)--Rutgers University, 2008. / "Graduate Program in Psychology." Includes bibliographical references (p. 33-36).

Thesis (Ph. D.)--Rutgers University, 2008. / "Graduate Program in Psychology." Includes bibliographical references (p. 83-91).

Thesis (Psy.D.)--Fairleigh Dickinson University, 2004. / Source: Dissertation Abstracts International, Volume: 65-09, Section: B, page: 4859. Chair: Ronald Dumont. Available also in print.

Thesis (Ph.D.)--Fairleigh Dickinson University, 2004. / Source: Dissertation Abstracts International, Volume: 65-02, Section: B, page: 1043. Chair: Neil A. Massoth. Available also in print.

<p> The majority of computational theories of inductive processes in psychology derive from small-scale experiments with simple stimuli that are easy to represent. However, real-world stimuli are complex, hard to represent efficiently, and likely require very different cognitive strategies to cope with. Indeed, the difficulty of such tasks are part of what make humans so impressive, yet methodological resources for modeling their solutions are limited. This presents a fundamental challenge to the precision of psychology as a science, especially if traditional laboratory methods fail to generalize. Recently, a number of computationally tractable, data-driven methods such as deep neural networks have emerged in machine learning for deriving useful representations of complex perceptual stimuli, but they are explicitly optimized in service to engineering objectives rather than modeling human cognition. It has remained unclear to what extent engineering models, while often state-of-the-art in terms of human-level task performance, can be leveraged to model, predict, and understand humans.</p><p> In the following, I outline a methodology by which psychological research can confidently leverage representations learned by deep neural networks to model and predict complex human behavior, potentially extending the scope of the field. In Chapter 1, I discuss the challenges to ecological validity in the laboratory that may be partially circumvented by technological advances and trends in machine learning, and weigh the advantages and disadvantages of bootstrapping from largely uninterpretable models. In Chapter 2, I contrast methods from psychology and machine learning for representing complex stimuli like images. Chapter 3 provides a first case study of applying deep neural networks to predict whether objects in a large database of images will be remembered by humans. Chapter 4 provides the central argument for using representations from deep neural networks as proxies for human psychological representations in general. To do this, I establish and demonstrate methods for quantifying their correspondence, improving their correspondence with minimal cost, and applying the result to the modeling of downstream cognitive processes. Building on this, Chapter 5 develops a method for modeling human subjective probability over deep representations in order to capture multimodal mental visual concepts such as "landscape". Finally, in Chapter 6, I discuss the implications of the overall paradigm espoused in the current work, along with the most crucial challenges ahead and potential ways forward. The overall endeavor is almost certainly a stepping stone to methods that may look very different in the near future, as the gains in leveraging machine learning methods are consolidated and made more interpretable/useful. The hope is that a synergy can be formed between the two fields, each bootstrapping and learning from the other.</p><p>

<p> Visual perception can be understood as an inferential process that combines noisy sensory information with internalized knowledge drawn from previous experience. In statistical Bayesian terms, internal representations of the visual environment can be understood as posterior estimates obtained by weighting imperfect sensory information (a likelihood) by internalized biases (a prior). Given limited perceptual resources, it is advantageous for the visual system to capitalize on predictable regularities of the visual world, and internalize them in the form of priors. This dissertation presents novel findings in the domain of spatial vision and visual memory, as well as some new work on memory for the 3D orientation of objects. In all cases, an unprecedented signal-to-noise ratio, achieved by employing serial reproduction chains (a &ldquo;telephone game&rdquo; procedure) combined with non-parametric kernel density estimation techniques, reveals a number of stunning intricacies in the prior for the first time. Methodological implications, as well as implications for amending prior empirical findings and revisiting past theoretical explanations are discussed.</p><p>

<p> Visual search has been extensively studied in the laboratory, yielding broad insights into how we search through and attend to the world around us. In order to know if these insights are valid, however, this research must not be confined to the sanitized imagery typically found within the lab. Comparatively little research has been conducted on visual search within naturalistic settings, and this gap must therefore be bridged in order to further our understanding of visual search. Based on the results of Experiment 1, Experiment 2 was conducted to test three common effects observed in previous studies of visual search: the effects of background complexity, target-background similarity, and target-distractor similarity on response time. Results show that these hypotheses carry over to the natural world, but also that there are other effects present not accounted for by current theories of visual search. The argument is made for the modification of these theories to incorporate this naturalistic information. </p><p>