This thesis examined the role of procedural learning in human probabilistic category learning (PCL). It was proposed that there was a lack of clear behavioural evidence for learning without awareness in PCL. Eleven experiments are reported that investigated the characteristics of learning in a prototypical probabilistic category learning task (the weather prediction task). The results were interpreted as contradicting the popular interpretation of weather prediction task learning as procedurally based. Rather, it was shown that behavioural data was consistent with declarative learning. This learning was not dissociable with measures of cue knowledge. Strategy analysis converged with the behavioural data, suggesting the dominance of declarative learning in this task. It was proposed that a single system account (e.g., Lagnado et al., 2006; Newell et al., 2007) which does not posit a role for procedural learning was the most appropriate way to understand learning in the weather prediction task.

This thesis examined the role of procedural learning in human probabilistic category learning (PCL). It was proposed that there was a lack of clear behavioural evidence for learning without awareness in PCL. Eleven experiments are reported that investigated the characteristics of learning in a prototypical probabilistic category learning task (the weather prediction task). The results were interpreted as contradicting the popular interpretation of weather prediction task learning as procedurally based. Rather, it was shown that behavioural data was consistent with declarative learning. This learning was not dissociable with measures of cue knowledge. Strategy analysis converged with the behavioural data, suggesting the dominance of declarative learning in this task. It was proposed that a single system account (e.g., Lagnado et al., 2006; Newell et al., 2007) which does not posit a role for procedural learning was the most appropriate way to understand learning in the weather prediction task.

Many tasks can easily be posed as the problem of responding to the states of an external world with actions that maximise the reward received over time. Algorithms that reliably solve such problems exist. However, their worst-case complexities are typically more than proportional to the size of the state space in which a task is to be performed. Many simple tasks involve enormous numbers of states, which can make the application of such algorithms impractical. This thesis examines reinforcement learning algorithms which effectively learn to perform tasks by constructing mappings from states to suitable actions. In problems involving large numbers of states, these algorithms usually must construct approximate, rather than exact, solutions and the primary issue examined in the thesis is the way in which the complexity of constructing adequate approximations scales as the size of a state space increases. The vast majority of reinforcement learning algorithms operate by constructing estimates of the long-term value of states and using these estimates to select actions. The potential effects of errors in such estimates are examined and shown to be severe. Empirical results are presented which suggest that minor errors are likely to result in significant losses in many problems, and where such losses are most likely to occur. The complexity of constructing estimates accurate enough to prevent significant losses is also examined empirically and shown to be substantial.

Thesis (Ph. D.)--State University of New York at Binghamton, Computer Science Department, 2006. / Includes bibliographical references.

Dissertation (Ph.D.) -- The Institute for Clinical Social Work, 1992. / A dissertation submitted to the faculty of the Institute of Clinical Social Work in partial fulfillment for the degree of Doctor of Philosophy.

Zugl.: Zugl.: Duisburg, Essen, Univ., Diss., 2006

Univ., Diplomarbeit, 2004--Frankfurt (Main)

Univ., Diplomarbeit, 2006--Frankfurt (Main)

Univ., Diplomarbeit, 2006--Frankfurt (Main)

Dissertation (Ph. D.)--University of Akron, Dept. of Curricular and Instructional Studies-Secondary Education, 2006. / "December, 2006." Title from electronic dissertation title page (viewed 04/28/2008) Advisor, Qetler Jensrud; Committee members, Sandra C. Coyner, Dennis Doverspike, Xin Liang, Carole Newman, Susan Olson; Interim Department Chair, Bridgie Ford; Dean of the College, Patricia Nelson; Dean of the Graduate School, George R. Newkome. Includes bibliographical references.