The present study was designed to analyze the transfer of stimulus control in temporal fading procedures. Several aspects of temporal fading procedures were manipulated including sources of inhibitory stimulus control, delays of reinforcement, and rates of increase in the temporal parameter of a fading procedure. In Experiment I, previous research producing transfer of stimulus control in a temporal fading procedure was directly replicated and controls were implemented for the operation of inhibition. The results showed that inhibitory stimulus control is not necessary in order to produce a transfer as participants with neutral stimulus backgrounds also transferred from one dimension to another without errors. However, positive stimulus backgrounds in the fading procedure prohibited the participants from achieving an errorless transfer of discrimination learning. In Experiment II, a fixed trial duration was employed with a constant and equal delay of reinforcement for both new and original stimulus dimensions. In this condition, participants did not transfer from one dimension to another with up to 30-second delays. Control participants were yoked to participants exposed to delayed and fading procedures to examine response latencies under delayed reinforcement for a simultaneous discrimination. There were no discernible response patterns under this condition except that participants continued to emit relatively short response latencies with a 40-second delay of reinforcement. In Experiment III, the effects of different steps of temporal fading on transfer were examined. The results showed that as the step of delay increased (10 sec. per trial), subjects transferred earlier in the fading series. Also, subjects with extremely low steps of delay (.1 sec. per trial) tended to remain with the original stimulus dimension. Experiments I through III demonstrated the necessity of either inhibitory or neutral stimulus backgrounds, differential delays of reinforcement correlated with each stimulus dimension, and relatively rapid increments in delay of the original stimulus dimension to obtain transfers of stimulus control in temporal fading procedures. When excitatory stimulus backgrounds were employed, or no differential delay of reinforcement was present, or the delay of the original stimulus dimension increased slowly, errorless transfers were not obtained. Overall, the results indicate that temporal fading procedures are a reliable, although complexly controlled, means of obtaining transfer between two stimulus dimensions.

Submitted by Fernanda Puga Santos null (ferpuga@gmail.com) on 2016-04-05T19:55:10Z No. of bitstreams: 1 Tese_FINAL.pdf: 3778155 bytes, checksum: 12218385281bca2455fe3172e45e514c (MD5) / Approved for entry into archive by Felipe Augusto Arakaki (arakaki@reitoria.unesp.br) on 2016-04-07T16:09:03Z (GMT) No. of bitstreams: 1 carvalho_fps_dr_prud.pdf: 3778155 bytes, checksum: 12218385281bca2455fe3172e45e514c (MD5) / Made available in DSpace on 2016-04-07T16:09:03Z (GMT). No. of bitstreams: 1 carvalho_fps_dr_prud.pdf: 3778155 bytes, checksum: 12218385281bca2455fe3172e45e514c (MD5) Previous issue date: 2016-03-31 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / Slope streaks são rastros escuros que se estendem por declives íngremes na superfície de Marte. Estes rastros representam um dos poucos processos geológicos ativos na superfície deste planeta. Atualmente, muitos pesquisadores os têm estudado com a finalidade de descobrir sua natureza, a qual permanece controversa. Além disso, os slope streaks clareiam com o tempo, fornecendo pistas sobre a deposição de poeira e também sobre a natureza do material da superfície. Embora exista um número considerável de pesquisadores que estudam esses rastros, a identificação destes ainda é realizada por especialistas manualmente, através de amostras de pequena dimensão. A existência de um número elevado destas estruturas na superfície de Marte, a necessidade de caracterizá-las e também de quantificar a sua evolução temporal, não pode continuar a ser efetuada simplesmente por amostragem e de uma forma manual. É neste contexto que esta pesquisa se enquadra. A proposta consiste em contribuir para a automação do processo de extração de informações em imagens da superfície de Marte, especificamente, extração de informações sobre slope streaks. Através do desenvolvimento de um método de detecção automática de slope streaks em imagens orbitais e, também, de um método automático para análise temporal da taxa de esmaecimento, este objetivo foi alcançado neste trabalho. O método de detecção desenvolvido baseia-se principalmente em Morfologia Matemática e faz uso de operadores morfológicos conectados para o pré-processamento das imagens, transformada top-hat, binarização pelo método de Otsu, afinamento e reconstrução geodésica, seguido por um filtro de fator de forma. O método para a análise temporal desenvolvido consiste em um algoritmo que calcula a taxa de contraste entre o interior e a área de vizinhança de um mesmo rastro, em imagens multi-temporais registradas. Os resultados obtidos com ambos os métodos foram bastante satisfatórios e possibilitaram extrair informações inovadoras a respeito do comportamento destes rastros na superfície de Marte. As duas ferramentas desenvolvidas mostraram-se robustas para serem aplicadas a grande escala e a um grande conjunto de imagens. / Slope streaks are typically dark, narrow and fan-shaped features that extend down slope on Mars surface. They are one of the most active and dynamic process observed on the planet’s surface. Dry and wet processes have been suggested for causing their formation but their origin is still unclear. Moreover, the streaks tend to fade with time, providing clues about dust settling and material properties. Studies that quantify some characteristics of these streaks are based on manual interactive procedures to delineate only a small portion of the available slope streaks and to measure their morphometric characteristics. The availability of a methodology to segment the streaks and to extract meaningful information would naturally increase the regional knowledge and the statistical significance, as a much larger amount of images from different locations could be analyzed, together with a more complete monitoring of the fading/appearance of the dark streaks. Thus, the purpose of this research is to contribute to the information extraction process from surface images of Mars. Hence, a method to automatically detect slope streaks and an algorithm to quantify the temporal fading of each streak over the years were developed. The detection method was based on morphological operators and it was used in a part of the methodology to quantify the fading of the streaks. The results of the detection method and the temporal fading algorithm were very satisfactory. Both methods are able to extract information about the behaviour of the slope streaks on the Martian surface. Finally, the two tools are robust enough to be applied on a large scale and a large set of images. / CAPES: 12791-13-0