Matching-to-sample learning and transfer with patterned visual stimuli in rats
17. Matching-to-sample learning and transfer with patterned visual stimuli in rats
University of Tokyo, Kazuo Okanoya-The University of Tokyo
Makiko Kamijo-The University of Tokyo, Kazuo Okanoya-The University of Tokyo
Abstract concepts, rules about relations between stimuli, have been argued to be unique to limited species and the set size of training stimuli is important to form the abstract concept. For example, pigeons learn stimulus-response chains on matching-to-sample (MTS) tasks with a small set size of stimuli whereas they show abstract concept learning as the set size is increased. Here, we examined whether rats could learn the MTS task with visual stimuli and transfer their learning to novel stimuli. Four Long-Evans rats were trained on MTS procedure with patterned black-and-white visual stimuli in an operant chamber attached with an LCD. A trial began with the presentation of a sample, and response to the sample resulted in the appearance of the two comparisons. A response to the identical stimulus with the sample was reinforced with pellets. The acquisition training was started with one stimulus-set (sample: A, comparisons: A and B), and another stimulus-set (sample: C, comparisons: C and D) was added after rats accomplished the learning criteria. When rats reached the criteria with two stimulus sets, they were tested with novel stimuli and repeated the tests as training stimuli were increased. All rats accomplished the criteria and three rats showed the significant transfer to novel stimuli after the training with two or four stimulus sets in test sessions. And the accuracy of two rats out of three was equivalent to the training trials. These results suggest rats could acquire the abstract MTS concept. Supported by JSPS grant #4903, 17H06380.
Perceptual Learning: discrimination based on different rules
18. Human Perceptual Learning: discrimination based on different rules
Marcela Lugo Hernández y Rosalva Cabrera Castañón
Facultad de Estudios Superiores Iztacala-UNAM
Marcela Lugo Hernández y Rosalva Cabrera Castañón Facultad de Estudios Superiores Iztacala-UNAM
The pre-exposition with compound stimuli had demonstrated that humans participants can to do more accurate discriminations in later tests as function of physical properties of the stimuli (Wang, Lavis, Hall & Mithell, 2008; Hall & Mitchell, 2014); nevertheless, it hasn´t demonstrated the rules abstracction with only to observe perceptual arrays that involved it. The purpose of this study is to evaluate in humans the perceptual learning of relations with compound visual stimuli. University students (n = 10) were randomly assigned to four groups. The stimuli, big white square (X) with a small square into (red or blue as distinctive characteristics A or B). During pre-exposition stimuli were arranged to exemplify two rules, which involved Colour and Position. Experimental group (EUD), was pre-exposed to arrays color 1-up position vs color 2-down-position; Experimental group (ELR) was pre-exposed to colour 1-left position vs colour 2-right position; the Control groups (CUD and CLR) were no pre-exposed. Later, in the test the four groups were exposed to choice multiple task in which they had to select three from six items, three of them (correct) exemplified the pre-exposed colour-position relation. The mean percent of correct responses was better to experimental groups (72%) that the control groups (45%), additionally, experimental groups showed an accuracy of 73% in their first choice. These results allow us to conclude that this experimental preparation of Perceptual Learning was effective to promoting the colour-position relationships learning. Key words: perceptive learning, rules, discrimination, colour-position, abstraction.